File: faq.html

package info (click to toggle)
msql 2.0.3-5
links: PTS
area: non-free
in suites: hamm, slink
size: 3,596 kB
ctags: 2,483
sloc: ansic: 34,769; sh: 1,309; yacc: 1,084; perl: 328; makefile: 267
file content (10408 lines) | stat: -rw-r--r-- 308,378 bytes
<html>
<head>
<title>The mSQL FAQ</title>
</head>
 
<body bgcolor="#ccccff">
<center>
<h1><a name="top">The mSQL FAQ</a></h1>
</center>
 
<h4><a name="rcs-revision">$Revision: 1.1 $</a></h4>
 
<h4><a name="rcs-date">$Date: 1998/02/10 21:18:45 $</a></h4>
 
<center>
<h2><a name="preamble">Preamble</a></h2>
</center>
 
A text copy of this FAQ can be obtained via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/faq.txt">ftp://bond.edu.au/pub/Minerva/msql/faq.txt</a>
</pre>
 
or an HTML copy can be obtained via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/faq.html">ftp://bond.edu.au/pub/Minerva/msql/faq.html</a>
</pre>
 
This FAQ is maintained by
<a href="http://www.uniq.com.au/people/peters.html">Peter Samuel</a>
<a href="mailto:Peter.Samuel@uniq.com.au">
&lt;Peter.Samuel@uniq.com.au&gt;</a> and is produced independently of
Hughes Technologies (the developers of mSQL). <p>
 
While every attempt is made to ensure that the information contained in
this FAQ is accurate, no guarantees of accuracy can or will be made. <p>
 
This FAQ attempts to deal with mSQL. Because of time and space
constraints, other applications, including value added applications
provided with the mSQL distribution such as W3-mSQL and Lite, will not
be covered in any great detail. <p>
 
<a href="#other-apps">Third party applications</a> mentioned in this
FAQ may not be compatible with the current release of mSQL - by
necessity their development will lag that of mSQL. If you have any
questions concerning their status please contact the
<a href="#mailing-list">mSQL mailing list</a> or the author of the
application in question. <p>
 
New questions in the FAQ are marked with <b>(=)</b>. Questions that
have been modified since the last release of the FAQ are marked with
<b>(-)</b>. <p>
 
<b>Note: BEFORE POSTING A QUESTION TO THE <a href="#mailing-list">mSQL
MAILING LIST</a>, PLEASE READ THE SECTION <a href="#posting"> &quot;How
do I post a question to the mSQL mailing list&quot;</a>.</b> <p>
 
<hr>
<hr width="75%">
 
<center>
<h2><a name="contents">Contents</a></h2>
</center>
 
<ul>
 
<li><b><a href="#general">General Information</a></b>
    <ol>
 
    <li> (=) <a href="#faq-covers">What is covered by this FAQ?</a>
 
    <li> <a href="#what-is-sql">What is SQL?</a>
 
    <li> <a href="#sqlbooks">Books on SQL and mSQL?</a>
 
    <li> <a href="#what-is-msql">What is mSQL?</a>
 
    <li> <a href="#platforms">What platforms will it run on?</a>
 
    <li> <a href="#shareware">Is it free?</a>
 
    <li> <a href="#where-is">Where can I get it?</a>
 
    <li> (-) <a href="#mirrors">Are there any mirror sites for
	mSQL?</a>
 
    <li> <a href="#current-version">What is the latest version?</a>
 
    <li> <a href="#next-version">When will the next version be
	released?</a>
 
    <li> (=) <a href="#beta-software">What is Beta-release
	software?</a>
 
    <li> <a href="#why-bambi">Why is David called Bambi?</a>
 
    <li> <a href="#prepared">How is this FAQ prepared?</a>
 
    <li> <a href="#helping-hand">How can I contribute to the FAQ?</a>
 
    </ol>
    <p>
 
<li><b><a href="#help-and-support">Help and Support</a></b>
 
    <ol>
 
    <li> <a href="#recommended-patches">Are there any recommended
	patches to the latest version?</a>
 
    <li> <a href="#known-bugs">Are there any known bugs or problems
	with the current release of mSQL?</a>
 
    <li> (-) <a href="#version2">What are the current problems with
	mSQL 2.0B<i>x</i>?</a>
 
    <li> <a href="#other-patches">Are there any other
	patches available?</a>
 
    <li> <a href="#how-to-patch">How do I apply patches to the mSQL
	source?</a>
 
    <li> <a href="#online-doco">Is online documentation
	available?</a>
 
    <li> <a href="#mailing-list">Is there a mailing list for
	mSQL?</a>
 
    <li> <a href="#mailing-list-digest">Is there a mailing list digest
	for mSQL?</a>
 
    <li> (-) <a href="#html-mail">Are there any HTML archives of the mSQL
	mailing list?</a>
 
    <li> <a href="#news-group">Is there a news group for mSQL?</a>
 
    <li> <b><font size="+1"><a href="#posting">How do I post a
	question to the mSQL mailing list? (PLEASE READ
	THIS!)</a></font></b>
 
    <li> <a href="#current-bug">I've found a bug. How can I tell if
	it's been fixed yet?</a>
 
    <li> <a href="#reporting-bugs">How do I report a bug in mSQL?</a>
 
    <li> <a href="#contrib-patches">How do I contribute a patch to
	mSQL?</a>
 
    <li> <a href="#contrib-code">How do I contribute code to the mSQL
	community?</a>
 
    <li> <a href="#code-examples">Where can I find examples of code
	that interface with mSQL?</a>
 
    <li> <a href="#porting">Are there any non Unix ports of
	mSQL?</a>
 
    </ol>
    <p>
 
<li><b><a href="#features-and-limits">Features and Limitations</a></b>
 
    <ol>
 
    <li> <a href="#support-progs">What support programs are bundled
	with mSQL?</a>
 
    <li> <a href="#sql-commands">What SQL commands are
	supported?</a>
 
    <li> <a href="#keys">Does mSQL support keys?</a>
 
    <li> <a href="#datatypes">What datatypes are supported?</a>
 
    <li> <a href="#api">What functions does the mSQL API provide?</a>
 
    <li> <a href="#views">Are views or virtual tables supported?</a>
 
    <li> <a href="#table-aliasing">Does mSQL support table
	aliasing?</a>
 
    <li> <a href="#column-constraints">Are column constraints
	supported?</a>
 
    <li> <a href="#stored-procedures">Are stored procedures
	supported?</a>
 
    <li> <a href="#acl">Are access privileges supported?</a>
 
    <li> <a href="#blobs">Does mSQL support BLOBs?</a>
 
    <li> <a href="#transactions">Are the transaction commands, BEGIN,
	COMMIT, and ROLLBACK supported?</a>
 
    <li> <a href="#table-limits">What are the limits on table and
	field names?</a>
 
    <li> <a href="#user-limits">What other limits can be modified?</a>
 
    <li> <a href="#db-size">How much data can mSQL address?</a>
 
    <li> <a href="#expressions">Are there any limitations in the way
	mSQL handles logical expressions?</a>
 
    <li> <a href="#return-values">How does mSQL return values?</a>
 
    <li> <a href="#return-order">How does SELECT return rows?</a>
 
    <li> <a href="#nested-tables">Can mSQL nest tables?</a>
 
    <li> <a href="#storage-overhead">What storage overheads does mSQL
	have?</a>
 
    <li> <a href="#ram">Does <tt>msqld</tt> allocate more RAM to itself
	as new databases are added?</a>
 
    <li> <a href="#perform">Does performance degrade as the number of
	databases increase?</a>
 
    <li> <a href="#cursors">Does mSQL support cursors?</a>
 
    <li> <a href="#intl">Does mSQL support different character
	sets?</a>
 
    <li> <a href="#afs">Why doesn't mSQL work when installed in
	disk space served by the Andrew File System (AFS)?</a>
 
    </ol>
    <p>
 
<li><b><a href="#install-problems">Installation Problems</a></b>
 
    <ol>
 
    <li> <a href="#irix-username">Under IRIX (SGI) I get problems
	relating to my username</a>
 
    <li> <a href="#osf1-hpux">On OSF/1 or HP-UX I have trouble starting
	<tt>msqld</tt> at boot time</a>
 
    <li> <a href="#osf1-gcc">Should I use <tt>cc</tt> or <tt>gcc</tt>
	when building mSQL on my Dec Alpha running OSF/1?</a>
 
    <li> <a href="#linux-mmap">Does mSQL work with Linux
	<tt>mmap()</tt>?</a>
 
    <li> <a href="#hpux-compile">Does mSQL work with HP-UX?</a>
 
    <li> <a href="#hpux-perl">I'm having trouble compiling
	MsqlPerl with mSQL under HP-UX</a>
 
    <li> <a href="#sco-unix">How can I install mSQL on a SCO Unix
	system?</a>
 
    <li> <a href="#linux-setup">Why does <tt>setup</tt> fail when
	building mSQL on a Linux system?</a>
 
    <li> <a href="#solaris">Why won't my third party applications
	compile under Solaris 2.<i>x</i>?</a>
 
    <li> <a href="#linux-setup">Why can't I run <tt>setup</tt>
	under Linux?</a>
 
    <li> <a href="#bind49">Why do I get errors about
	&quot;<tt>sys/bitypes.h</tt>&quot; when compiling under Solaris
	2.5?</a>
 
    <li> <a href="#makedepend">Why do I get errors about
	&quot;<tt>POSIX_C_SOURCE</tt>&quot; when compiling under
	Solaris 2.5?</a>
 
    <li> <a href="#sgi-libs">Why do I get errors about
	&quot;<tt>weak definitions</tt>&quot; when compiling under
	IRIX?</a>
 
    <li> <a href="#aix-ps2">How can I build mSQL version 1.0.16 for
	an AIX system running on an IBM PS/2?</a>
 
    <li> <a href="#apache-auth">How can I make <tt>w3-auth.c</tt>
	work with my Apache HTTPD server?</a>
 
    </ol>
    <p>
 
<li><b><a href="#runtime-problems">Runtime Problems</a></b>
 
    <ol>
 
    <li> <a href="#root-user"><tt>msqladmin</tt> will not let me create
	a database</a>
 
    <li> <a href="#acl-file">When I start <tt>msqld</tt> it complains
	about an ACL file</a>
 
    <li> <a href="#pid-file">When I start <tt>msqld</tt> it complains
	about a PID file</a>
 
    <li> <a href="#recompile">I've just installed the latest version of
	mSQL and now my own applications won't work!</a>
 
    <li> <a href="#setuid">Access control doesn't work with my
	setuid applications</a>
 
    <li> <a href="#address-in-use">Why do I see an &quot;<tt>Address
	already in use</tt>&quot; error message when I attempt to start
	<tt>msqld</tt>?</a>
 
    <li> <a href="#bind-permissions">Why do I see a &quot;<tt>Can't
	start server : UNIX Bind : Permission denied</tt>&quot; error
	message when I attempt to start <tt>msqld</tt>?</a>
 
    <li> <a href="#tmpdir">Why do I see &quot;<tt>Couldn't create
	temporary table</tt>&quot; error messages?</a>
 
    <li> <a href="#tmp-space">How can I avoid running out of space
	when doing certain complex table joins?</a>
 
    <li> <a href="#bzero-missing"><tt>msqld</tt> is suddenly
	dumping core and complaining about <tt>bzero()</tt></a>
 
    <li> <a href="#bsdlib">Why does <tt>relshow</tt> drop the first
	two characters from its output?</a>
 
    <li> <a href="#msqladmin">Why won't <tt>msqladmin</tt>
	work?</a>
 
    <li> <a href="#sgi-reals">Why does <tt>msqld</tt> crash when
	attempting to display real numbers on 64-bit SGI platforms?</a>
 
    <li> <a href="#aix-compress">Why won't mSQL work with the
	compressed file system under AIX?</a>
 
    <li> <a href="#hpux-index">Why do I see &quot;<tt>Protocol
	mismatch"</tt> errors under HP-UX?</a>
 
    <li> <a href="#machten">Why do I see <tt>&quot;Can't start
	server :  UNIX Bind : Invalid argument&quot;</tt> errors
	running <tt>msqld</tt> under a MachTen BSD Unix system?</a>
 
    <li> <a href="#socket-location">Where did my mSQL socket file
	go?</a>
 
    <li> <a href="#slow-connections">Why does it take such a long
	time to connect to the mSQL database server?</a>
 
    <li> <a href="#beta-expire">Why do I see <tt>&quot;This Beta
	test release expires in 57 days&quot;</tt> when I start
	<tt>msql2d</tt>?</a>
 
    <li> <a href="#acl-ips">Can I use IP addresses in the mSQL ACL
	file?</a>
 
    <li> <a href="#protocol-mismatch">What does the &quot;Protocol
	mismatch. Server Version = 22 Client Version = 6&quot; error
	mean?</a>
 
    <li> <a href="#query-ordering">Why does mSQL take longer to
	return the same results when I change the order of my
	query?</a>
 
    <li> <a href="#multi-table-selects">Why don't my SELECT
	statements work across multiple tables?</a>
 
    </ol>
    <p>
 
<li><b><a href="#how-do-i">How do I ....?</a></b>
 
    <ol>
 
    <li> <a href="#single-quotes">How do I embed single quotes in a
	field?</a>
 
    <li> <a href="#escape-chars">What other characters need special
	treatment?</a>
 
    <li> <a href="#null-fields">How do I handle null fields?</a>
 
    <li> <a href="#no-case">How do I perform case insensitive
	matches?</a>
 
    <li> <a href="#add-column">How do I add a column to an existing
	table?</a>
 
    <li> <a href="#parent-child">When should I call
	<tt>msqlConnect()</tt> in a parent/child situation?</a>
 
    <li> <a href="#reserved-words">Can I use mSQL reserved words as
	field or table names?</a>
 
    <li> <a href="#maxmin">How do I find the maximum or minimum value
	in a table?</a>
 
    <li> <a href="#schema">How can I determine the structure of a
	database?</a>
 
    <li> <a href="#reconnect">What happens when the mSQL server
	goes down between requests?</a>
 
    <li> <a href="#muli-servers">Can I run more than one copy of
	<tt>msqld</tt> on the same CPU?</a>
 
    <li> <a href="#unique-keys">How can I automatically ensure that
	each record receives a unique primary key?</a>
 
    <li> <a href="#multiple-includes">How can I avoid compiler
	redefinition errors when compiling my own mSQL applications?</a>
 
    <li> <a href="#linking">How do I link the mSQL library with my own
	code?</a>
 
    <li> <a href="#numrows">How can I find the number of rows in a
	table?</a>
 
    <li> <a href="#delete-all">How can I delete all the rows in a table
	without destroying the table structure?</a>
 
    <li> <a href="#ascii-import">How can I import flat files or
	other database formats into an mSQL table?</a>
 
    <li> <a href="#touched">How can I determine the number of rows
	affected by my last query?</a>
 
    <li> <a href="#w3auth">How can I get <tt>w3-msql</tt> to work
	with <tt>w3-auth</tt>?</a>
 
    </ol>
    <p>
 
<li> (-) <b><a href="#other-apps">Contributed Code and Third Party
    Applications</a></b> <p>
 
<li> (-) <b><a href="#web-sites">Web sites using mSQL</a></b>
 
</ul>
 
<hr>
<hr width="75%">
 
<center>
<h2><a name="general">General Information</a></h2>
</center>
 
<h3><a name="faq-covers">What is covered by this FAQ?</a></h3>
 
This FAQ covers mSQL itself. It does not attempt to answer questions
about various value added packages such as <a href="#w3-msql">W3-mSQL</a>,
<a href="#perl">MsqlPerl</a>, <a href="#tcl">msqltcl</a>,
<a href="#php">PHP/FI</a>, etc in any great detail. Nor does it attempt
to address problems integrating HTML and mSQL in anything but a cursory
manner. <p>
 
The reasons for this limited scope are simple - time and space. I
maintain the FAQ out of the goodness/madness of my heart. I simply
don't have the time to cover everything - if I did cover everything I'd
never get any real work done and I'd starve :( <p>
 
The FAQ is also quite large - 297k as of release 1.32 and growing. If I
could include minute details about all the value added mSQL products
you'd be looking at a very large document indeed (or several large
documents). This is the sort of problem I don't have the time to deal
with :) <p>
 
There is a bright side however. I'm perfectly willing to accept
contributions to the FAQ on <b>ANY</b> mSQL related issue - in fact I
highly encourage it. If you have details concerning an area of mSQL
that you'd like included in the FAQ, write it up and send it to me
<a href="mailto:Peter.Samuel@uniq.com.au">
&lt;Peter.Samuel@uniq.com.au&gt;</a>. <p>
 
Some of the value added products have mailing lists and home pages.
See the <a href="#other-apps">&quot;Contributed Code and Third Party
Applications&quot;</a> section for details. <p>
 
<hr>
 
<h3><a name="what-is-sql">What is SQL?</a></h3>
 
The following section is based on chapter 1 of &quot;Oracle7 Server -
SQL Language Reference Manual&quot; - Oracle Corporation, December
1992, part number 778-70-1292. <p>
 
<b>SQL</b> is an acronym that stands for <b>S</b>tructured <b>Q</b>uery
<b>L</b>anguage. It is often pronounced &quot;<i>sequel</i>&quot;. It
was developed in the mid 1970s by IBM. <p>
 
The American National Standards Institute (ANSI) and the International
Standards Organisation (ISO) have adopted SQL as the standard language
for relational database management systems. <p>
 
SQL provides commands for a variety of tasks including: <p>
 
<ul>
    <li> querying data
    <li> inserting, updating and deleting rows in a table
    <li> creating, replacing, altering and dropping objects
    <li> controlling access to the database and its objects
    <li> guaranteeing database consistency and integrity
</ul>
 
SQL provides easy to learn commands that are both consistent and
applicable for all users. <p>
 
While most relational database management systems - including mSQL -
provide support for SQL, each vendor usually has their own unique
extensions to the language that may hinder the portability of SQL
procedures from one database platform to another. <p>
 
JCC Consulting <a href="http://www.jcc.com">http://www.jcc.com</a>
maintains a central source of information about the SQL standards
process and its current state. For more details visit: <p>
 
<pre>
    <a href="http://www.jcc.com/sql_stnd.html">http://www.jcc.com/sql_stnd.html</a>
</pre>
 
<hr>
 
<h3><a name="sqlbooks">Books on SQL and mSQL</a></h3>
 
There are countless books available on database design and SQL. Some
good starting points (in no particular order) are: <p>
 
<pre>
<font size="-1">
    C. J. Date
    &quot;An Introduction to Database Systems&quot
    Vol I, Sixth Edition, 1995
    Addison Wesley
    ISBN 0-201-54329-X
    (<a href="http://heg-school.aw.com/cseng/authors/date/intro/intro.html">http://heg-school.aw.com/cseng/authors/date/intro/intro.html</a>)
 
    C. J. Date and Hugh Darwen
    &quot;A Guide to Sql Standard&quot;
    Fourth Edition, 1997
    Addison Wesley
    ISBN 0-201-96426-0
    (<a href="http://heg-school.aw.com/cseng/authors/date/sql/sql.html">http://heg-school.aw.com/cseng/authors/date/sql/sql.html</a>)
 
    Judith S. Bowman, Sandra L. Emerson and Marcy Darnovsky
    &quot;The Practical SQL Handbook: Using Structured Query Language&quot;
    Third Edition
    Addison Wesley
    ISBN 0-201-44787-8
    (<a href="http://www.awl.com/devpress/titles/44787.html">http://www.awl.com/devpress/titles/44787.html</a>)
 
    Terry Halpin
    &quot;Conceptual Schema and Relational Database Design&quot;
    1995
    Prentice Hall
    ISBN 0-13-355702-2
    (<a href="http://www.prenhall.com/allbooks/ptr_0133557022.html">http://www.prenhall.com/allbooks/ptr_0133557022.html</a>)
 
    Joe Celko
    &quot;Instant SQL Programming&quot;
    1995
    Wrox Press
    ISBN 1-874416-508
    (<a href="http://www.wrox.com/scripts/bookdetail.idc?Code=508">http://www.wrox.com/scripts/bookdetail.idc?Code=508</a>)
 
    Martin Gruber
    &quot;Understanding SQL&quot;
    1990
    Sybex Publishing
    ISBN  0-89588-644-8
    (<a href="http://www.sybex.com/cgi-bin/bookpg.pl?644back.html">http://www.sybex.com/cgi-bin/bookpg.pl?644back.html</a>)
 
    Jeff Rowe
    &quot;Building Internet Database Servers with CGI&quot;
    1996
    New Riders Publishing
    ISBN 1-56205-573-9
    (<a href="http://merchant.superlibrary.com:8000/catalog/hg/PRODUCT/PAGE/15620/bud/1562055739.html">http://merchant.superlibrary.com:8000/catalog/hg/PRODUCT/PAGE/15620/bud/1562055739.html</a>)
 
    Perkins &amp; Morgan
    &quot;Teach yourself SQL in 14 days&quot;
    1995
    SAMS Publishing
    ISBN 0-67230-855-X
    (<a href="http://www.connection.se/cgi-bin/shop/method=showitem/sid=bm/uid=78F52F3/item=067230855X">http://www.connection.se/cgi-bin/shop/method=showitem/sid=bm/uid=78F52F3/item=067230855X</a>
    or
    <a href="http://merchant.superlibrary.com:8000/catalog/hg/PRODUCT/PAGE/06723/bud/067230855X.html">http://merchant.superlibrary.com:8000/catalog/hg/PRODUCT/PAGE/06723/bud/067230855X.html</a>)
 
    Connolly, Begg &amp; Strachan
    &quot;Database Systems: A Practical Approach to Design,
	Implementation and Management&quot;
    1996
    Addison Wesley
    ISBN 0-201-42277-8
    (<a href="http://heg-school.aw.com/cseng/authors/connolly/databsys/databsys.html">http://heg-school.aw.com/cseng/authors/connolly/databsys/databsys.html</a>)
 
    Melton &amp; Simon
    &quot;Understanding the New SQL: A Complete Guide&quot;
    1993
    Morgan Kaufmann
    ISBN 1-55860-245-3
 
    Mike Morgan &amp; Jeff Wandling
    &quot;Webmaster Expert Solutions&quot;
    1996
    QUE books
    ISBN 0-7897-0801-9
    (<a href="http://merchant.superlibrary.com:8000/catalog/hg/PRODUCT/PAGE/07897/bud/0789708019.html">http://merchant.superlibrary.com:8000/catalog/hg/PRODUCT/PAGE/07897/bud/0789708019.html</a>)
    (see <a href="http://www.speakeasy.org/~jdw">http://www.speakeasy.org/~jdw</a> for errata)
 
    Ramez Elmasri and Shamkant B. Navathe
    &quot;Fundamentals of Database Systems&quot;
    Second Edition
    Addison Wesley
    ISBN 0-8053-1748-1
    (<a href="http://heg-school.aw.com/cseng/authors/elmasri/Dbase2e/Dbase2e.html">http://heg-school.aw.com/cseng/authors/elmasri/Dbase2e/Dbase2e.html</a>)
</font>
</pre>
 
<hr>
 
<h3><a name="what-is-msql">What is mSQL?</a></h3>
 
mSQL is mini SQL, a light weight database engine developed by
<a href="http://Bond.edu.au/Bond/Admin/ITS/People/bambi">
David J. Hughes</a>
<a href="mailto:bambi@hughes.com.au">&lt;bambi@hughes.com.au&gt;</a> at
<a href="http://bond.edu.au">Bond University</a>, Australia. It has
been designed to provide fast access to stored data with low memory
requirements. As its name implies mSQL offers a subset of SQL as its
query interface. Although it only supports a subset of SQL, everything
it supports is in accordance with the ANSI SQL specification. <p>
 
<hr>
 
<h3><a name="platforms">What platforms will it run on?</a></h3>
 
mSQL has been developed under Sun OS 4.1.1 but has been tested under
Solaris 2.<i>x</i> (release 2.3, 2.4 and 2.5), Ultrix 4.3, Linux, and
OSF/1 (cc not gcc). That said, it should &quot;<tt>autoconf</tt>&quot;
and build on most BSD derived systems, SVR4 based systems or POSIX
O/S's (that should cover most of them). It has been reported that it
works out-of-the-box on HP-UX, NeXT, SCO, Sequent, Cray, Tandem, *BSD
and a few others. <p>
 
<hr>
 
<h3><a name="shareware">Is it free?</a></h3>
 
David Hughes writes: <p>
 
<pre>
    &quot;mSQL has been released in the past under terms known as
    'conscience-ware', the basic concept of which was that companies
    that used the software could contribute a small amount to the
    continued development of the software without any strict rules
    being placed upon such 'donations'. Although the concept sounds
    fair, it failed badly with only 3 contributions being made from
    over 3,600 copies of mSQL-1.0.5 that were ftp'ed from my machine
    alone. Over 1,000 of those copies went to commercial organisations
    and I receive many questions a day from companies using mSQL behind
    their WWW servers etc who are looking for free support. <p>
 
    In an attempt to balance this out and allow me to devote some time
    to mSQL (rather than other pursuits that I do to generate an
    income), mSQL is now shareware. I still believe in free software
    over the Internet and cooperation in research so the new license is
    designed not to hurt Universities, research groups and other people
    that _should_ have free access to software. Commercial
    organisations that are using this so that they don't have to buy an
    Oracle or an Ingres will now have to buy mSQL (at a minute fraction
    of the cost of one of the commercial offerings). <p>
 
    Please read the doc/License file to see if you are required to
    register your copy. An invoice is included in both Postscript and
    ASCII format to ease the generation of payments.&quot;
</pre>
 
You may freely use mSQL if and only if you fall into the categories
outlined in the mSQL License file: <p>
 
<pre>
    You can use this software free of charge if you are an educational
    institution (excluding commercial training organisations),
    non-commercial research organisation, registered charity,
    registered not-for-profit organisation, or full-time student.
</pre>
 
If you do not fall into any of these categories, you will have to pay
a license fee to use mSQL. As of release 1.0.16, the cost of mSQL is:
<p>
 
<pre>
    Commercial Installation - AUD $225
 
    Private Installation - AUD $65
</pre>
 
Exchange rates may vary wildly, but at the time of preparing this FAQ,
the Australian dollar was trading at about 0.75 US dollars. This
information is provided as an indication only. You MUST check your
local exchange rates before preparing to purchase mSQL. <p>
 
An online currency conversion system developed by
<a href="http://www.olsen.ch">Olsen &amp; Associates</a> is available
at <a href="http://www.olsen.ch/cgi-bin/exmenu">
http://www.olsen.ch/cgi-bin/exmenu</a>. <p>
 
Another online currency conversion system is available from
<a href="http://www.DynaMind-LLC.com/services/utilities/currency.cgi">
http://www.DynaMind-LLC.com/services/utilities/currency.cgi</a>. <p>
 
<hr>
 
<h3><a name="where-is">Where can I get it?</a></h3>
 
mSQL 1.<i>x</i> can be obtained via anonymous ftp from
<a href="http://bond.edu.au">Bond University</a>, Australia. The latest
version of release 1.<i>x</i> can be found in the directory: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/">ftp://bond.edu.au/pub/Minerva/msql/</a>
</pre>
 
The latest beta version of 2.<i>x</i> can be found at: <p>
 
<pre>
    <a href="http://hughes.com.au/software/msql2/current.htm">http://hughes.com.au/software/msql2/current.htm</a>
</pre>
 
<hr>
 
<h3><a name="mirrors">Are there any mirror sites for mSQL?</a></h3>
 
If you would like to mirror the mSQL FAQ, please contact
<a href="mailto:Peter.Samuel@uniq.com.au">
&lt;Peter.Samuel@uniq.com.au&gt;</a>. <p>
 
The following sites mirror the mSQL FAQ: <p>
 
<ul>
 
    <li> <a href="http://www.swl.fh-heilbronn.de/msql">
    http://www.swl.fh-heilbronn.de/msql</a> <p>
 
    Thanks to Tobias Haecker
    <a href="mailto:tobi@swl.fh-heilbronn.de">
    &lt;tobi@swl.fh-heilbronn.de&gt;</a>. <p>
 
    <li> <a href="ftp://ftp.nerosworld.com/pub/SQL/msql">
    ftp://ftp.nerosworld.com/pub/SQL/msql</a> <p>
 
    Thanks to David Perry <a href="mailto:deperry@nerosworld.com">
    &lt;deperry@nerosworld.com&gt;</a>. <p>
 
    <li> <a href="http://cal038031.student.utwente.nl/~mSQL">
    http://cal038031.student.utwente.nl/~mSQL</a> or
    <a href="ftp://cal038031.student.utwente.nl/pub/mSQL">
    ftp://cal038031.student.utwente.nl/pub/mSQL</a> <p>
 
    Thanks to Sjoerd de Heer
    <a href="mailto:sjoerd@cal038031.student.utwente.nl">
    &lt;sjoerd@cal038031.student.utwente.nl&gt;</a>. <p>
 
    <li> <a href="http://sunsite.srce.hr/mirror/mSQL-FAQ/faq.html">
    http://sunsite.srce.hr/mirror/mSQL-FAQ/faq.html</a> <p>
 
    Thanks to Davorin Bengez
    <a href="mailto:Davorin.Bengez@sunsite.srce.hr">
    &lt;Davorin.Bengez@sunsite.srce.hr&gt;</a>. <p>
 
    <li>
    <a href="http://www.gold-cousins.org/lite_site/ftp/faq.html">
    http://www.gold-cousins.org/lite_site/ftp/faq.html</a> <p>
 
    Thanks to Jeanne Gold <a href="mailto:goldsaic@ix.netcom.com">
    &lt;goldsaic@ix.netcom.com&gt;</a>. <p>
 
    <li>
    <a href="http://tcts.fpms.ac.be/~zanoni/msql/faq.html">
    http://tcts.fpms.ac.be/~zanoni/msql/faq.html</a> <p>
 
    Thanks to Laurent Zanoni
    <a href="mailto:zanoni@tcts.fpms.ac.be">
    &lt;zanoni@tcts.fpms.ac.be&gt;</a>. <p>
 
    <li>
    <a href="http://audi.dglnet.com.br/saito/mirrors/msql/faq.html">
    http://audi.dglnet.com.br/saito/mirrors/msql/faq.html</a> and <br>
    <a href="http://audi.dglnet.com.br/saito/mirrors/msql/faq.txt">
    http://audi.dglnet.com.br/saito/mirrors/msql/faq.txt</a> <p>
 
    Thanks to Ricardo Saito <a href="mailto:rtsaito@dglnet.com.br">
    &lt;rtsaito@dglnet.com.br&gt;</a>. <p>
 
    <li>
    <a href="http://www.blnet.com/msqlpc/lozano/msqlfaq.htm">
    http://www.blnet.com/msqlpc/lozano/msqlfaq.htm</a>. <p>
 
    Thanks to Fernando Lozano <a href="mailto:bl@blnet.com">
    &lt;bl@blnet.com&gt;</a>.
 
</ul>
 
The following sites mirror the mSQL 1.<i>x</i> software distribution:
<p>
 
<ul>
    <li> <a href="ftp://ftp.ntua.gr/pub/databases/msql/">
    ftp://ftp.ntua.gr/pub/databases/msql/</a> <p>
 
    Thanks to Yiorgos Adamopoulos
    <a href="mailto:adamo@noc.ntua.gr">&lt;adamo@noc.ntua.gr&gt;</a>.
    <p>
 
    <li> <a href="ftp://sunsite.auc.dk/pub/databases/msql">
    ftp://sunsite.auc.dk/pub/databases/msql</a> <p>
 
    Thanks to Jesper Hagen <a href="mailto:hagen@iesd.auc.dk">
    &lt;hagen@iesd.auc.dk&gt;</a> <p>
 
    <li> <a href="ftp://ftp.nerosworld.com/pub/SQL/msql">
    ftp://ftp.nerosworld.com/pub/SQL/msql</a> <p>
 
    Thanks to David Perry <a href="mailto:deperry@nerosworld.com">
    &lt;deperry@nerosworld.com&gt;</a>. <p>
 
    <li>
    <a href="ftp://sunsite.srce.hr/mirrors/bond.edu.au/pub/Minerva/msql/msql">
    ftp://sunsite.srce.hr/mirrors/bond.edu.au/pub/Minerva/msql/msql</a>
    <p>
 
    Thanks to Davorin Bengez
    <a href="mailto:Davorin.Bengez@sunsite.srce.hr">
    &lt;Davorin.Bengez@sunsite.srce.hr&gt;</a>. <p>
 
    <li> <a href="ftp://ftp.bf.rmit.edu.au/pub/msql">
    ftp://ftp.bf.rmit.edu.au/pub/msql</a> <p>
 
    Thanks to Malcolm Herbert
    <a href="mailto:mjch@boojum.bf.rmit.edu.au">
    &lt;mjch@boojum.bf.rmit.edu.au&gt;</a>.
 
</ul>
<p>
 
The following sites mirror the mSQL 2.<i>x</i> software distribution:
 
<ul>
 
    <li> <a href="ftp://ftp.worldone.com/pub/msql">
    ftp://ftp.worldone.com/pub/msql</a> <p>
 
    Thanks to WorldOne <a href="http://www.worldone.com">
    http://www.worldone.com</a>. <p>
 
    <li> <a href="ftp://ftp.nerosworld.com/pub/msql/msql2">
    ftp://ftp.nerosworld.com/pub/msql/msql2</a> <p>
 
    Thanks to David Perry <a href="mailto:deperry@nerosworld.com">
    &lt;deperry@nerosworld.com&gt;</a>. <p>
 
    <li> <a href="http://cal038031.student.utwente.nl/~mSQL">
    http://cal038031.student.utwente.nl/~mSQL</a> or
    <a href="ftp://cal038031.student.utwente.nl/pub/mSQL">
    ftp://cal038031.student.utwente.nl/pub/mSQL</a> <p>
 
    Thanks to Sjoerd de Heer
    <a href="mailto:sjoerd@cal038031.student.utwente.nl">
    &lt;sjoerd@cal038031.student.utwente.nl&gt;</a>. <p>
 
    <li>
    <a href="ftp://sunsite.srce.hr/mirrors/bond.edu.au/pub/Minerva/msql/msql2">
    ftp://sunsite.srce.hr/mirrors/bond.edu.au/pub/Minerva/msql/msql2</a>
    <p>
 
    Thanks to Davorin Bengez
    <a href="mailto:Davorin.Bengez@sunsite.srce.hr">
    &lt;Davorin.Bengez@sunsite.srce.hr&gt;</a>.
</ul>
 
The following sites have a complete mirror of
<a href="http://hughes.com.au">http://hughes.com.au</a>: <p>
 
<ul>
    <li>
    <a href="http://hermes.nerosworld.com/msql">
    http://hermes.nerosworld.com/msql</a> <p>
 
    Thanks to David Perry <a href="mailto:deperry@nerosworld.com">
    &lt;deperry@nerosworld.com&gt;</a>.
 
</ul>
<p>
 
<hr>
 
<h3><a name="current-version">What is the latest version?</a></h3>
 
At the time of compiling this FAQ, the latest released version was
1.0.16. <p>
 
It is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/msql-1.0.16.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/msql-1.0.16.tar.gz</a>
    (195705 bytes)
</pre>
 
<hr>
 
<h3><a name="next-version">When will the next version be
released?</a></h3>
 
David has suspended all development work on mSQL version 1.<i>x</i> and
is concentrating his efforts on version 2.<i>x</i>. However, he will
continue to provide any necessary bug fixes for the current release of
mSQL. <p>
 
mSQL development is NOT David's primary role. He does this work in
whatever spare time he has available so PLEASE don't pester him with
requests about when we can expect future releases. Take the zen
approach and just let it happen :) <p>
 
A beta snapshot of mSQL version 2.0.<i>x</i> was released on May 8
1997. It is available from: <p>
 
<pre>
    <a href="http://hughes.com.au/software/msql2/msql-2.0-B7.1.tar.gz">http://hughes.com.au/software/msql2/msql-2.0-B7.1.tar.gz</a>
    (373261 bytes)
</pre>
 
<b>Note:</b> Do NOT download the mSQL-2.0B7.0 distribution. If you do
so you will not be able to compile the Lite modules. Use the
mSQL-2.0B7.1 distribution instead. <p>
 
It is also mirrored at a number of other sites. See the section above
titled &quot;<a href="#mirrors">Are there any mirror sites for
mSQL?</a>&quot; for details on these mirror sites. <p>
 
More details on mSQL 2.0.<i>x</i> can be found by visiting the Hughes
Technologies web pages at
<a href="http://hughes.com.au">http://hughes.com.au</a>. <p>
 
<hr>
 
<h3><a name="beta-software">What is Beta-release software?</a></h3>
 
Bambi is very interested in making mSQL work as well as possible for as
many people as possible. To make that happen quickly, Bambi has released
versions of mSQL which are not fully tested, with the hope that people
will try them under more circumstances than he could ever have the
time or equipment to do himself. <p>
 
This means, however, that this Beta software <b>MAY NOT WORK
CORRECTLY</b> for you. If you cannot tolerate hangs, crashes, or loss
of data, you should use the <a href="#current-version">latest released
version</a> and any <a href="#recommended-patches"> recommended
patches</a> (currently mSQL version 1.0.16 with the &quot;lost table
patch&quot;). However, if you can use the new version, please do so and
report bugs to the <a href="#mailing-list">mailing list</a> and to the
<a href="#reporting-bugs">bugs list</a> so they can be fixed in the
next release. <p>
 
<hr>
 
<h3><a name="why-bambi">Why is David called Bambi?</a></h3>
 
Bambi is a nickname that David acquired quite a long time ago. He
claims there are no derogatory implications associated with the name.
If you'd like to find out more send him some mail at
<a href="mailto:bambi@hughes.com.au">&lt;bambi@hughes.com.au&gt;</a>.
<p>
 
<hr>
 
<h3><a name="prepared">How is this FAQ prepared?</a></h3>
 
The FAQ is written in HTML and proofread using version 3.01Gold of the
Netscape Navigator running under Solaris 2.5.1. The HTML syntax is
checked using version 1.017 of Neil Bowers'
<a href="mailto:neilb@khoral.com">&lt;neilb@khoral.com&gt;</a> weblint
package. Text versions of the FAQ are created by saving the document as
a text file from within the Netscape Navigator. <p>
 
(See <a href="http://www.cre.canon.co.uk/~neilb/weblint.html">
http://www.cre.canon.co.uk/~neilb/weblint.html</a> for more details on
weblint). <p>
 
<hr>
 
<h3><a name="helping-hand">How can I contribute to the FAQ?</a></h3>
 
If you find any errors or omissions in the FAQ, <b>please</b> contact
me, Peter Samuel <a href="mailto:Peter.Samuel@uniq.com.au">
&lt;Peter.Samuel@uniq.com.au&gt;</a>. I will be happy to correct any
errors or add elements to the FAQ. <p>
 
If you want something added to the FAQ, it would be a great help if you
could write a section and forward it to me - then I can simply insert
it in the appropriate location. <p>
 
Remember, I don't own the FAQ - I'm just its caretaker on your behalf.
<p>
 
<hr>
<hr width="75%">
 
<center>
<h2><a name="help-and-support">Help and Support</a></h2>
</center>
 
<h3><a name="recommended-patches">Are there any recommended patches to
the latest version?</a></h3>
 
There are no official patches to any release of mSQL. If modifications
need to be made, a new version of mSQL will be released. <p>
 
The above not withstanding there is one unofficial patch that should
be applied to mSQL version 1.0.16. <p>
 
<a name="p1">This</a> patch comes from David Hughes
<a href="mailto:bambi@hughes.com.au">&lt;bambi@hughes.com.au&gt;</a>
and fixes the very old and very obscure problem of the occasionally
disappearing table. <p>
 
David writes:
 
<pre>
    After sitting here banging my head against the old &quot;missing table&quot;
    bug with my partner in crime Jason <a href="mailto:jason@fan.net.au"> &lt;jason@fan.net.au&gt;</a> , we've fixed
    it !!!!!  This is yet another very obscure bug.  So, can it be
    reproduced?  Yup, once you know the problem.
 
	o Fill the table cache
	o Cause a reference to a table that doesn't exist
	o The oldest cache entry will have the table definition list
	  set to NULL but with the name, DB, and age fields still set
 
    We all owe a debt of thanks to Jason as it was his inability to type
    table names correctly that found the problem :)
</pre>
 
<pre>
*** ./src/msql/msqldb.c.orig    Sun Jul  7 16:47:09 1996
--- ./src/msql/msqldb.c Tue Jul 30 17:07:42 1996
***************
*** 1398,1403 ****
--- 1398,1406 ----
                safeFree(entry-&gt;rowBuf);
                safeFree(entry-&gt;keyBuf);
                entry-&gt;def = NULL;
+               *(entry-&gt;DB) = 0;
+               *(entry-&gt;table) = 0;
+               entry-&gt;age = 0;
        }
 
        /*
</pre>
 
To avoid conflicts with mismatched whitespace,
<a href="#how-to-patch">apply</a> this patch using the
&quot;<tt>-l</tt>&quot; option of <tt>patch</tt>.
(<i>That's an &quot;el&quot;, NOT a &quot;one&quot; or an
&quot;eye&quot;</i>). <p>
 
David assures us that this patch will be rolled into version 1.0.17. <p>
 
<hr>
 
<h3><a name="known-bugs">Are there any known bugs or problems with the
current release of mSQL?</a></h3>
 
Since the release of mSQL 1.0.12, several problems have been reported
to the <a href="#mailing-list">mSQL mailing list</a>. These are outlined
below. <p>
 
<ul>
    <li> <b>Problems with table/field name lengths</b> <p>
 
    The definition of
    <tt><b><a href="#table-limits">NAME_LEN</a></b></tt> changed from
    version 1.0.10 to 1.0.12. It now seems that 18 characters is the
    maximum value for a table or field name. This problem exists in
    mSQL versions 1.0.12 and above. <p>
 
    David Hughes writes: <p>
 
<pre>
    The reason it changed in 1.0.12 is because I had a couple of
    field_name buffers that didn't have room for a trailing NULL. Now,
    trying to fix that without altering the size of the struct implies
    that you have to shorten the name of the field (to leave the
    required room). I didn't want to force everyone out there to drop
    and reload every database they have just because of a 1 byte buffer
    over-run.
 
    If this is a major problem for you then ....
 
	o dump all your databases
	o find the definition in question
	o set it to a value you like
	o rebuild everything
	o reload everything
 
    If you can work out what to change to get the extra field name
    length then you know enough about what you are doing to do the rest
    of what's required. I haven't provided a step-by-step because if
    people can't find the value they have to change I'm sure they'd
    still end up asking the list about this stuff after breaking
    things.
</pre>
 
    At present there is NO patch to fix this problem. Possible
    workarounds are David's method outlined above or to downgrade to
    version 1.0.10. <p>
 
    <li> The regular expression code in version 1.0.14 and above has
    been cleaned up with the unfortunate side affect of making it
    impossible to search for the special characters <tt>^</tt> and
    <tt>\</tt>.  <p>
 
    The cleanup also changes the way in which other special characters
    can be searched. The tables in the section
    <a href="#escape-chars">&quot;What other characters need
    special treatment?&quot;</a> outline the differences for versions
    up to and including version 1.0.13 and 1.0.14 and above. <p>
 
    <li> There are still isolated reports of <tt>msqld</tt> crashing
    with SIGSEGV (segmentation violation) errors under Linux. These
    occurrences have diminished but have not disappeared. There has
    been one report that the problem lies not within mSQL or Linux but
    in the hardware in use. An upgrade of the hardware may eliminate
    the SIGSEGV conditions. For more non mSQL specific details on this
    problem see: <p>
 
<pre>
    <a href="http://www.bitwizard.nl/sig11">http://www.bitwizard.nl/sig11</a>
</pre>
 
    <li> Some users have reported the <tt>msqld</tt> consumes large
    amounts of memory when doing ORDERed selects: <p>
 
    Marty Picco writes:
 
<pre>
    I have a 200K+ record database running on a p90 with 96MB memory.
    I have noticed that the first SELECT I do on this database after
    msqld is started causes msqld to grab as much memory as it can...in
    this case about 68MB. It appears that the memory is never released
    until the daemon is restarted. Indeed, the SELECT does have an
    ORDER clause.
</pre>
 
    David has said that he'll investigate this problem. <p>
 
    <li> There are still isolated reports of mSQL behaving strangely
    under HP-UX 9.<i>x</i>. Any strangeness is more than likely a
    symptom of a broken <tt>mmap()</tt> implementation. Try
    compiling mSQL without <tt>mmap()</tt> support by editing <p>
 
<pre>
    ./targets/<i>your-architecture</i>/site.mm
</pre>
 
    and ensuring the <tt>mmap()</tt> directive reads: <p>
 
<pre>
    MMAP=
</pre>
 
    and then compile and re-install the mSQL package.
 
</ul>
<p>
 
<hr>
 
<h3><a name="version2">What are the current problems with mSQL
2.0B<i>x</i>?</a></h3>
 
<b>Note:</b> This section is temporary. It will be merged into the
rest of the FAQ as mSQL 2.<i>x</i> matures. <p>
 
<b>Note:</b> Some parts of this section have been compiled by Lloyd
Parkes <a href="mailto:Lloyd.Parkes@vuw.ac.nz">
&lt;Lloyd.Parkes@vuw.ac.nz&gt;</a>. <p>
 
There are a few known problems with the current release of mSQL
2.0B<i>x</i>. This list is undoubtedly incomplete and will probably
remain so because of the &quot;moving target&quot; nature of new
software. <p>
 
<ul>
    <li> mSQL 2.<i>x</i> requires a fully functioning <tt>mmap()</tt>.
    If your operating system does not support <tt>mmap()</tt> then you
    can't use mSQL 2.<i>x</i>. The following operating systems are
    known to provide insufficient <tt>mmap()</tt> support: <p>
 
	<ul>
	    <li> NeXT <p>
 
	    <li> Linux - kernel 1.2.<i>x</i> and below. (Linux
	    1.3.<i>x</i> and above should be fine). <p>
 
	    <li> A/UX 3.1.1 <p>
 
	    <li> DEC Ultrix <p>
 
	    <li> MachTen <p>
 
	    <li> UNICOS (Cray research)
	</ul>
	<p>
 
    <li> There are some new keywords which cannot be used as table or
    field names. See the table in the section <a href="#reserved-words">
    &quot;Can I use mSQL reserved words as field or table
    names?&quot;</a> below. <p>
 
    <li> The internal format of the database files has changed
    significantly. If you wish to use mSQL 1.<i>x</i> databases with
    mSQL 2.<i>x</i> you must reload them. The following procedure
    should be followed: <p>
 
    <ol>
	<li> Using mSQL 1.<i>x</i>, dump your database with
	<tt>msqldump</tt>: <p>
 
<pre>
    msqldump -h <i>dbhost</i> -c <i>dbname</i> &gt; db.dump
</pre>
 
	<li> Edit <tt>db.dump</tt> and change ALL occurrences of
	<tt>PRIMARY KEY</tt> definitions to <tt>INDEX</tt>. <p>
 
	<li> Edit <tt>db.dump</tt> and change ALL occurrences of mSQL
	keywords used as table or fields names. <p>
 
	<li> Using mSQL 2.<i>x</i>, reload your database with
	<tt>msql</tt>: <p>
 
<pre>
    msql -h <i>dbhost</i> <i>dbname</i> &lt; db.dump
</pre>
 
    </ol>
 
    Vivek Khera <a href="mailto:khera@kci.kciLink.com">
    &lt;khera@kci.kciLink.com&gt;</a> has provided the following Perl
    program that will help you convert your mSQL 1.<i>x</i> dumps to
    the mSQL 2.<i>x</i> format: <p>
 
<pre>
<font size="-1">
    #! /usr/local/bin/perl
    use strict;
 
    # read a msql 1.x dump and convert PRIMARY KEY entries to UNIQUE INDEX
    # called pkey for that table.
    #
    # also converts fields named time-&gt;timestamp, date-&gt;datestamp, count-&gt;counter
 
    # V. Khera (vivek@khera.org) 08-JAN-1996
 
    # Usage: msqldumpconvert orig.dumpfile &gt; new.dumpfile
    # or use it in a pipeline between msqldump from 1.x and msql from 2.x
 
 
    my($field, $table);
    while (&lt;&gt;) {
      if (m/^\# mSQL Dump/) {
	print &quot;# mSQL Dump converted to mSQL 2.0 format\n&quot;;
      } elsif (m/^\s+ (\w+).*PRIMARY KEY,/) {
	$field = $1;
	s/ PRIMARY KEY//;
	print;
      } elsif (m/CREATE TABLE (\w+)/) {
	$table = $1;
	$field = &quot;&quot;;		# no field means no primary key
	print;
      } elsif (m/^\) \\g/ and $field) {
	print;
	print &quot;\nCREATE UNIQUE INDEX pkey ON $table ($field)\\g\n\n&quot;;
      } elsif (m/^  count /) {
	s/^  count /  counter /;
	print;
	print STDERR &quot;Converted field 'count' to 'counter' in table $table\n&quot;;
      } elsif (m/^  time /) {
	s/^  time /  timestamp /;
	print;
	print STDERR &quot;Converted field 'time' to 'timestamp' in table $table\n&quot;;
      } elsif (m/^  date /) {
	s/^  date /  datestamp /;
	print;
	print STDERR &quot;Converted field 'date' to 'datestamp' in table $table\n&quot;;
      } else {
	print;
      }
    }
</font>
</pre>
 
    <li> The mSQL 2.<i>x</i> API has some new functionality. You MUST
    re-link all your third party applications with the new API. This
    should be generally successful as the new API is a superset of the
    mSQL 1.<i>x</i> API. <p>
 
    <b>Note:</b> Simply relinking will NOT make the new functionality
    available - the third party application must take advantage of the
    new functions for you to make use of them. Relinking will only
    enable you to use your existing third party applications with the
    mSQL 2.<i>x</i> database server. <p>
 
    <li> mSQL 2.<i>x</i> uses a different protocol than version
    1.<i>x</i>. The tools provided with mSQL version 1.<i>x</i>, such
    as <tt>msql</tt>, <tt>msqladmin</tt>, <tt>relshow</tt>, etc will
    not work with mSQL version 2.<i>x</i>. You MUST use the tools
    provided with the version 2.<i>x</i> distribution. <p>
 
    <li> You cannot perform regular expression searches on
    <tt>TEXT</tt> fields. <p>
 
    David Hughes writes: <p>
 
<pre>
    TEXT fields will always be slower than CHAR fields because the data
    is split over multiple buffers. As for the searching you can do
    anything other than *LIKE testing at the moment. I'll try to work
    out a way to do LIKE and CLIKE on TEXT fields. RLIKE is a
    non-option.
</pre>
 
    <li> mSQL 2.0B5 uses a different database format than previous
    2.<i>x</i> releases. This change was made to accommodate 64 bit
    aligned data for machines running 64 pit processors such as the
    DEC Alpha. If you are upgrading from a previous 2.<i>x</i> release
    to 2.0B5 or beyond, you must dump your databases and reload them
    using the following procedure: <p>
 
	<ol>
	    <li> Use <tt>msqldump</tt> from the mSQL version
	    2.<i>x</i> release you are currently using to dump your
	    databases: <p>
 
<pre>
    msqldump -h <i>dbhost</i> -c <i>dbname</i> &gt; db.dump
</pre>
 
	    <li> Drop the old database using <tt>msqladmin</tt> from
	    the mSQL version 2.<i>x</i> release you are currently using
	    to dump your databases: <p>
 
<pre>
    msqladmin drop <i>dbname</i>
</pre>
 
	    <li> Remove the old version of mSQL 2.<i>x</i> and install
	    the new version which should be mSQL version 2.0B5 or
	    above. <p>
 
	    <li> Create the new database using <tt>msqladmin</tt> from
	    mSQL version 2.0B5: <p>
 
<pre>
    msqladmin create <i>dbname</i>
</pre>
 
	    <li> Use the <tt>msql</tt> monitor from mSQL 2.0B5 to
	    reload the dumped data: <p>
 
<pre>
    msql -h <i>dbhost</i> <i>dbname</i> &lt; db.dump
</pre>
 
    </ol>
 
    <li> <tt>msqlimport</tt> has a number of problems. <p>
 
    <ul>
	<li> Empty fields in the input are not parsed as NULL fields,
	or anything else for that matter.
 
	<li> There seems to be a problem with memory being trashed.
 
	<li> Lossage when a delimiter longer than one character is
	specified.
 
    </ul>
    <p>
 
    A patch is available from <a href="http://www.vuw.ac.nz/~lloyd/msql">
    http://www.vuw.ac.nz/~lloyd/msql</a>. <p>
 
    <li> <tt>msqldump</tt> has trouble dumping the new <tt>DATE</tt>
    type fields. It doesn't surround them with single quotes. This
    patch from Andreas Wadephul
    <a href="mailto:wadib@uhura.asfh-berlin.de">
    &lt;wadib@uhura.asfh-berlin.de&gt;</a> will fix this problem: <p>
 
<pre>
<font size="-1">
*** src/msql/msqldump.c.orig	Tue Mar 25 09:49:57 1997
--- src/msql/msqldump.c	Tue Mar 25 09:50:57 1997
***************
*** 346,352 ****
  			if (row[i])
  			{
  				if (field-&gt;type == CHAR_TYPE ||
! 				    field-&gt;type == TEXT_TYPE)
  				{
  					tmp = escapeText(row[i]);
  					printf(&quot;\'%s\'&quot;, tmp);
--- 346,353 ----
  			if (row[i])
  			{
  				if (field-&gt;type == CHAR_TYPE ||
! 				    field-&gt;type == TEXT_TYPE ||
! 				    field-&gt;type == DATE_TYPE)
  				{
  					tmp = escapeText(row[i]);
  					printf(&quot;\'%s\'&quot;, tmp);
</font>
</pre>
 
    To avoid conflicts with mismatched whitespace,
    <a href="#how-to-patch">apply</a> this patch using the
    &quot;<tt>-l</tt>&quot; option of <tt>patch</tt>.
    (<i>That's an &quot;el&quot;, NOT a &quot;one&quot; or an
    &quot;eye&quot;</i>). <p>
 
    <li> Heikki Vatiainen <a href="mailto:hessu@cs.tut.fi">
    &lt;hessu@cs.tut.fi&gt;</a> has found a problem where the server
    (or client?) gets very confused with <tt>INSERT</tt> queries that
    have large amounts of data. <p>
 
    More details are available from
    <a href="http://www.vuw.ac.nz/~lloyd/msql">
    http://www.vuw.ac.nz/~lloyd/msql</a>. <p>
 
    <li> Various errors with the bookmarks sample application. <p>
 
    <li> Eric Huss <a href="mailto:e-huss@uiuc.edu">
    &lt;e-huss@uiuc.edu&gt;</a> reports there seems to be some
    inconsistencies in the behaviour of the mSQL API calls
    <tt>msqlListFields()</tt> and <tt>msqlListIndex()</tt>.  When
    passing a table name to <tt>msqlListFields()</tt>, if there is a
    newline character at the end of the table name, there is no
    problem. However, when passing a table name to
    <tt>msqlListIndex()</tt>, and the table name has a newline
    character at the end, then it doesn't work. This is due to the fact
    that msql is using <tt>strok(NULL,"\n\r")</tt> to chop up the
    arguments, and the extra newline in the middle causes problems. <p>
 
    <li> Brendan Quinn <a href="mailto:brendan@nexus.sofcom.com.au">
    &lt;brendan@nexus.sofcom.com.au&gt;</a> reports a problem with
    regular expression searches. <p>
 
    Brendan writes: <p>
 
<pre>
    If you have a field that is full and you do a like/clike search
    where the last character of the like string matches the last
    character of the field, mSQL-2.0 b3 and 4 (at least) incorrectly
    match.
 
    Example
    -------
 
    mSQL &gt; create table blah ( Organisation CHAR(50) ) \g
    Query OK.  1 row(s) modified or retrieved.
    mSQL &gt; insert into blah Values ('1995 Hopman Cup Tennis Tournament, Perth
    Western A') \g
    Query OK.  1 row(s) modified or retrieved.
    mSQL &gt; select * from blah where Organisation clike '%abababa%' \g
    Query OK.  1 row(s) modified or retrieved.
    +----------------------------------------------------+
    | Organisation                                       |
    +----------------------------------------------------+
    | 1995 Hopman Cup Tennis Tournament, Perth Western A |
    +----------------------------------------------------+
 
    which of course is wrong.
</pre>
 
    Brendan also supplies a patch for this problem: <p>
 
<pre>
<font size="-1">
*** src/msql/types.c.orig	Fri Apr 11 08:58:30 1997
--- src/msql/types.c	Tue Apr 15 16:53:00 1997
***************
*** 138,144 ****
                          {
                                  if (*cp1 == *cp2)
                                  {
! 					if(likeTest(cp1,cp2,length-count)==1)
                                          {
                                                  return(1);
                                          }
--- 138,144 ----
                          {
                                  if (*cp1 == *cp2)
                                  {
!                                       if(likeTest(cp1,cp2,length-count+1)==1)
                                          {
                                                  return(1);
                                          }
***************
*** 224,230 ****
                          {
                                  if (toupper(*cp1) == toupper(*cp2))
                                  {
! 					if(cLikeTest(cp1,cp2,length-count)==1)
                                          {
                                                  return(1);
                                          }
--- 224,230 ----
                          {
                                  if (toupper(*cp1) == toupper(*cp2))
                                  {
!                                       if(cLikeTest(cp1,cp2,length-count+1)==1)
                                          {
                                                  return(1);
                                          }
</font>
</pre>
 
    To avoid conflicts with mismatched whitespace,
    <a href="#how-to-patch">apply</a> this patch using the
    &quot;<tt>-l</tt>&quot; option of <tt>patch</tt>.
    (<i>That's an &quot;el&quot;, NOT a &quot;one&quot; or an
    &quot;eye&quot;</i>). <p>
 
    <li> Robert Sum <a href="mailto:sumrn@crd.ge.com">
    &lt;sumrn@crd.ge.com&gt;</a> reports on a number of
    problems with pattern matching in mSQL version 2.0B<i>x</i>.
    He has provided patches for these problems. The patches are
    available via anonymous ftp from: <p>
 
 
<pre>
    <a href="ftp://ce-toolkit.crd.ge.com/pub/sumrn/msql-2.0-B7/src/msql">ftp://ce-toolkit.crd.ge.com/pub/sumrn/msql-2.0-B7/src/msql</a>
</pre>
 
    <li> mSQL version 2.0B7.1 has a problem with unique indices.
    Dennis Doherty <a href="mailto:djd@atrieva.com">
    &lt;djd@atrieva.com&gt;</a> has supplied the following patch: <p>
 
<pre>
*** src/msql/index.c.orig	Tue Jun 24 10:26:54 1997
--- src/msql/index.c	Tue Jun 24 10:27:29 1997
***************
*** 353,359 ****
	if (node)
	{
		curRow = (u_int)node-&gt;data;
! 		if (curRow != rowNum &amp;&amp; rowNum != NO_POS)
			return(0);
	}
	return(1);
--- 353,359 ----
	if (node)
	{
		curRow = (u_int)node-&gt;data;
!                if (curRow != rowNum || rowNum == NO_POS)
			return(0);
	}
	return(1);
</pre>
 
    To avoid conflicts with mismatched whitespace,
    <a href="#how-to-patch">apply</a> this patch using the
    &quot;<tt>-l</tt>&quot; option of <tt>patch</tt>.
    (<i>That's an &quot;el&quot;, NOT a &quot;one&quot; or an
    &quot;eye&quot;</i>).
 
</ul>
 
<hr>
 
<h3><a name="other-patches">Are there any other patches
available?</a></h3>
 
This FAQ only details those
<a href="#recommended-patches">recommended patches</a> that are
necessary for the general operation of mSQL. There are a number of
other user contributed patches that enhance the operation of mSQL.
Some of these patches are available via anonymous ftp while others are
available from the mSQL mailing list
<a href="#list-archives">archives</a>. <p>
 
Some of these patches are for older releases of mSQL. They may have
been rolled into subsequent releases of mSQL or they may not work with
later releases of mSQL or they may break the successful operation of
later releases of mSQL. Use them at your own risk. <p>
 
A brief and no doubt incomplete list of these patches follows: <p>
 
<ul>
    <li> <a name="insert-patch">When</a> you insert a NULL as a primary
    key integer field, mSQL chooses a unique sequence number.
    Contributed by Jens-Peter Haack
    <a href="mailto:peter@netcs.com">&lt;peter@netcs.com&gt;</a> <p>
 
    Available from: <p>
 
<pre>
    <a href="ftp://ftp.uni-bremen.de/pub/unix/database/Minerva/patches/insert.msql-1.0.9">ftp://ftp.uni-bremen.de/pub/unix/database/Minerva/patches/insert.msql-1.0.9</a>
    (593 bytes)
</pre>
 
    <li> Mathematical updates. You can use statements like: <p>
 
<pre>
    update account set balance = balance + 100 where nr = 12345;
</pre>
 
    Contributed by Michael Koehne
    <a href="mailto:kraehe@bakunin.north.de">
    &lt;kraehe@bakunin.north.de&gt;</a>. <p>
 
    Available from: <p>
 
<pre>
    <a href="ftp://ftp.uni-bremen.de/pub/unix/database/Minerva/patches/update.msql-1.0.9">ftp://ftp.uni-bremen.de/pub/unix/database/Minerva/patches/update.msql-1.0.9</a>
    (12677 bytes)
</pre>
 
    <li> You can now use double quotes where quotes are allowed.
    Contributed by Michael Koehne
    <a href="mailto:kraehe@bakunin.north.de">
    &lt;kraehe@bakunin.north.de&gt;</a>. <p>
 
    Available from: <p>
 
<pre>
    <a href="ftp://ftp.uni-bremen.de/pub/unix/database/Minerva/patches/quote.msql-1.0.9">ftp://ftp.uni-bremen.de/pub/unix/database/Minerva/patches/quote.msql-1.0.9</a>
    (1225 bytes)
</pre>
 
    <li> Michael Koehne <a href="mailto:kraehe@bakunin.north.de">
    &lt;kraehe@bakunin.north.de&gt;</a> has bundled his value added
    patches to mSQL, which include the two mentioned above as well as
    some others. <p>
 
    Michael writes: <p>
 
<pre>
    This patch includes the following features :
 
    - Usage of double quotes for strings. I know this is not ISO
      but a lot of other databases also allow double quotes and
      some of my programs use double quotes.
    - Speedup of regular expression.
    - Simple expressions in update.
    - auto primary key insert.
    - Some bug fixes.
</pre>
 
    They are available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.uni-bremen.de/pub/unix/database/Minerva/features-msql-1.0.16">ftp://ftp.uni-bremen.de/pub/unix/database/Minerva/features-msql-1.0.16</a>
    (19297 bytes)
</pre>
 
    or
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLpatches/features.msql-1.0.16">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLpatches/features.msql-1.0.16</a>
    (19297 bytes)
</pre>
 
    <li> Case insensitive searching. You can use statements like: <p>
 
<pre>
    select name from table where name ilike 'Manu%'
</pre>
 
    Contributed by Benjamin Jacquard
    <a href="mailto:jacquard@lix.polytechnique.fr">
    &lt;jacquard@lix.polytechnique.fr&gt;</a>. <p>
 
    The patch is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.geo.tu-freiberg.de/pub/unix/msql/case-insensitive.patch">ftp://ftp.geo.tu-freiberg.de/pub/unix/msql/case-insensitive.patch</a>
    (8265 bytes)
</pre>
 
 
    <li> Martin Schulze
    <a href="mailto:joey@finlandia.Infodrom.North.DE">
    &lt;joey@finlandia.Infodrom.North.DE&gt;</a> has modified
    <tt>msqldump</tt> to let it insert or remove fields while dumping.
    Using this patch fixes the problem where the table definition
    cannot be modified after filling it with data. <p>
 
    The patch is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.infodrom.north.de/pub/mirror/msql/Patches/msqldump-1.0.16">ftp://ftp.infodrom.north.de/pub/mirror/msql/Patches/msqldump-1.0.16</a>
    (5275 bytes)
</pre>
 
    <li> Denis Ballant <a href="mailto:brdbb@brus.wec.com">
    &lt;brdbb@brus.wec.com&gt;</a> has rewritten the <tt>acl.c</tt>
    file from the mSQL distribution so that it will accept the wild card
    characters * and ? anywhere within an entry. Details of this patch
    are available from the mSQL mailing list
    <a href="#list-archives">archives</a> for the month of August
    1996. <p>
 
    <li> Robert Sum <a href="mailto:sumrn@crd.ge.com">
    &lt;sumrn@crd.ge.com&gt;</a> has a patch for mSQL
    version 1.0.16 that will restore the <tt>LIKE</tt> operation to its
    former glory - that is you can now search for the characters
    <tt>\</tt> and <tt>^</tt>. <p>
 
    Robert writes:
 
<pre>
    1.  Adjusted the deletion of backslashes during the mSQL to regexp
	translator code to make treatment of special characters more
	uniform. In other words, all regexp special characters except
	backslash require only two backslashes in front of them to be
	treated normally. (One is stripped by msql or the translator;
	the second by the regexp package.) The backslash character
	requires four backslashes because it is the escape character
	for both the msql monitor and the regexp package. The single
	quote, of course, remains the same needing only one backslash
	because it is special only to mSQL.
 
	So, you can all now search for \ and ^ and everything else. Or
	so my ever expanding test set tells me.
 
    2.  The execution of some of the non-LIKE comparison code by the
	NOT LIKE operator is fixed. (This should probably be a
	miniscule performance increase--except that it might be eaten
	by 1.)
</pre>
 
<pre>
<font size="-1">
*** src/msql/msqldb.c	1996/11/04 23:13:54	1.1
--- src/msql/msqldb.c	1996/11/05 01:02:01	1.2
***************
*** 3164,3176 ****
  		switch(*cp1)
  		{
  			case '\\':
! 				if (*(cp1+1) == '%' || *(cp1+1) == '_')
! 				{
! 					cp1++;
! 					*cp2 = *cp1;
  				}
- 				cp1++;
- 				cp2++;
  				break;
 
  			case '_':
--- 3164,3209 ----
  		switch(*cp1)
  		{
  			case '\\':
! 				/* RNS: The only time that we really want to
! 				   drop backslashes is when they are escaping
! 				   either % or _.
! 				   */
! 				if (!*(cp1+1)) {
!
! 				  /* keep backslash at the end of an expr. */
! 				  *cp2 = *cp1;
! 				  *cp1++; *cp2++;
!
! 				} else {
!
! 				  if ((*(cp1+1) == '%') || (*(cp1+1) == '_')) {
!
! 				    /* drop backslash when followed by
! 				       % or _.
! 				       */
! 				    cp1++;
! 				    *cp2 = *cp1;
! 				    cp1++; cp2++;
!
! 				  } else if (*(cp1+1) == '\\') {
!
! 				    /* keep both backslashes when they escape
! 				       themselves
! 				       */
! 				    *cp2 = *cp1;
! 				    *cp1++; *cp2++;
! 				    *cp2 = *cp1;
! 				    *cp1++; *cp2++;
!
! 				  } else {
!
! 				    /* keep the backslash so it can escape
! 				       some other regexp character.
! 				       */
! 				    *cp2 = *cp1;
! 				    *cp1++; *cp2++;
! 				  }
  				}
  				break;
 
  			case '_':
***************
*** 3351,3357 ****
  	REG	char	*c1,*c2;
  	REG	int	offset;
 
! 	if (op != LIKE_OP)
  	{
  		c1 = v1;
  		c2 = v2;
--- 3384,3394 ----
  	REG	char	*c1,*c2;
  	REG	int	offset;
 
! 	if ((op == LIKE_OP) || (op == NOT_LIKE_OP))
! 	{
! 		cmp = regexpTest( v1, v2, maxLen );
! 	}
! 	else
  	{
  		c1 = v1;
  		c2 = v2;
***************
*** 3395,3405 ****
  			break;
 
  		case LIKE_OP:
! 			result = regexpTest(v1,v2,maxLen);
  			break;
 
  		case NOT_LIKE_OP:
! 			result = !(regexpTest(v1,v2,maxLen));
  			break;
  	}
  	return(result);
--- 3432,3442 ----
  			break;
 
  		case LIKE_OP:
! 			result = cmp;
  			break;
 
  		case NOT_LIKE_OP:
! 			result = !cmp;
  			break;
  	}
  	return(result);
</font>
</pre>
 
    <li> Vivek Khera <a href="mailto:khera@kci.kciLink.com">
    &lt;khera@kci.kciLink.com&gt;</a> has a patch for increasing the
    number of connections that can be queued by <tt>msqld</tt>. The
    patch is available from the mSQL mailing list
    <a href="#list-archives">archives</a> for the month of December 1996.
    <p>
 
    <li> Sascha Kettler
    <a href="mailto:kettler@rummelplatz.uni-mannheim.de">
    &lt;kettler@rummelplatz.uni-mannheim.de&gt;</a> has patched mSQL
    version 1.0.16 such that it supports the Secure Sockets Layer -
    SSL - protocol. The patch is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLSSL/msql-1.0.16-patch.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLSSL/msql-1.0.16-patch.tar.gz</a>
    (7183 bytes)
</pre>
 
    <li> Etienne Klein <a href="mailto:etienne@lca.u-nancy.fr">
    &lt;etienne@lca.u-nancy.fr&gt;</a> has provided an <tt>mmap()</tt>
    substitute for NeXTSTEP systems. This code was developed by Fabien
    Roy <a href="mailto:Fabien_Roy@free.fdn.fr">
    &lt;Fabien_Roy@free.fdn.fr&gt;</a> and Robert Ehrlich
    <a href="mailto:Robert.Ehrlich@inria.fr">
    &lt;Robert.Ehrlich@inria.fr&gt;</a>. It is presented here in full.
    <p>
 
    Etienne writes: <p>
 
<pre>
    I just want to share this little patch which solve the mmap problem
    for NeXTSTEP. It apparently works (meaning all msql tests are OK)
    for:
 
	NSFIP 3.3 p1
	NS Motorola 3.3 p1.
	It does NOT work for HP 3.2
 
    All other configurations are untested !
    Thanks to the authors Fabien Roy and Robert Ehrlich for this patch !
 
    Etienne Klein
    Laboratoire de Chimie Analytique
    Faculte de Pharmacie de Nancy
    France
</pre>
 
<pre>
<font size="-1">
/*
 * @(#)map.c 1.0 of 20 December 1996
 *
 *      Copyright (c) 1996 by Fabien Roy.
 *      Written by Fabien Roy and	Robert Ehrlich.
 *		Fabien_Roy@free.fdn.fr 		Robert.Ehrlich@inria.fr
 *		Not derived from licensed software.
 *
 *      Permission is granted to anyone to use this software for any
 *      purpose on any computer system, and to redistribute it freely,
 *      subject to the following restrictions:
 *
 *      1. The author is not responsible for the consequences of use of
 *              this software, no matter how awful, even if they arise
 *              from defects in it.
 *
 *      2. The origin of this software must not be misrepresented, either
 *              by explicit claim or by omission.
 *
 *      3. Altered versions must be plainly marked as such, and must not
 *              be misrepresented as being the original software.
 *
 */
#include &lt;sys/types.h&gt;
#include &lt;sys/mman.h&gt;
#include &lt;stdlib.h&gt;
#include &lt;syscall.h&gt;
 
 
caddr_t
mmap(caddr_t addr, size_t len, int prot, int flags, int fd, off_t off)
 
{
	int pagelessone = getpagesize() -1;
	int size;
	caddr_t pageaddress;
 
		/* round to next page size */
		size = (len + pagelessone) &amp; ~pagelessone;
		/* allocate aligned pages */
		if (!(pageaddress = (caddr_t) valloc(size)))
			return (caddr_t) -1;
		/* map it */
		if (syscall(SYS_mmap, pageaddress, size, prot,
flags, fd, off)){
			free(pageaddress);
			return (caddr_t) -1;
		}
		return pageaddress;
}
 
void
munmap(caddr_t addr, size_t len)
{
	syscall(SYS_munmap,addr,len);
	free(addr);
}
</font>
</pre>
 
 
</ul>
 
<hr>
 
<h3><a name="how-to-patch">How do I apply patches to the mSQL
source?</a></h3>
 
Patches are distributed as context based difference listings. That is
the line by line differences between the original file and the new file
are listed with surrounding lines of code to provide some context
information. They are usually <a href="#contrib-patches">generated</a>
by using a program such as <tt>diff</tt>. <p>
 
The easiest way to apply these patches is to use Larry Wall's
<tt>patch</tt> program: <p>
 
<pre>
    <i>Patch</i> will take a patch file containing any of the four forms of
    difference listing produced by the <i>diff</i> program and apply those
    differences to an original file, producing a patched version. By
    default, the patched version is put in place of the original, with
    the original file backed up to the same name with the extension
    &quot;.orig&quot;.
</pre>
 
<tt>Patch</tt> is available from a number of anonymous ftp sites
worldwide. Latest versions of <tt>patch</tt> are being distributed by
the Free Software Foundation as part of the GNU suite of products. <p>
 
If you're having difficulty finding the latest version of
<tt>patch</tt>, you can download version 2.1 via anonymous ftp from:
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Misc/patch-2.1.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Misc/patch-2.1.tar.gz</a>
    (74856 bytes)
</pre>
 
While some patches are fairly simple - often involving minor changes
to a single line of code - others are quite complex and attempting to
apply these patches by hand is definitely NOT recommended. Use the
<tt>patch</tt> program whenever you need to apply a patch. <p>
 
To apply the <a href="#recommended-patches">patches listed in this
FAQ</a>, use this procedure:
 
<ol>
    <li> Save the patch to a file - say <tt>/tmp/msql-patch1</tt>.
    Only save the patch information - the <tt>patch</tt> program can
    usually determine which parts of the file are patches and which
    parts are not but its easier for you to strip out any unnecessary
    information before running <tt>patch</tt>. <p>
 
    You should be left with a file containing a number of sections
    similar to: <p>
 
<pre>
*** ./src/msql/net.c.orig	Tue Oct 17 11:24:03 1995
--- ./src/msql/net.c	Tue Dec 12 15:24:11 1995
***************
*** 66,72 ****
 
 
 
! void writePkt(fd)
	int	fd;
  {
	u_char	*cp;
--- 66,72 ----
 
 
 
! int writePkt(fd)
	int	fd;
  {
	u_char	*cp;
</pre>
 
    <li> Change directories to your mSQL source repository. For
    example: <p>
 
<pre>
    cd /usr/local/src/db/Minerva/msql/msql-1.0.16
</pre>
 
    <li> Apply the patch as follows: <p>
 
<pre>
    patch -l &lt; /tmp/msql-patch1
</pre>
 
    The &quot;<tt>-l</tt>&quot; option is used to tell <tt>patch</tt>
    to ignore any whitespace mismatches between lines in the patch file
    and lines in the mSQL source file. (<i>That's an &quot;el&quot;,
    NOT a &quot;one&quot; or an &quot;eye&quot;</i>). <p>
 
    <tt>Patch</tt> will respond with output similar to: <p>
 
<pre>
    Hmm...  Looks like a new-style context diff to me...
    The text leading up to this was:
    --------------------------
    |*** ./src/msql/net.c.orig        Tue Oct 17 11:24:03 1995
    |--- ./src/msql/net.c     Tue Dec 12 15:24:11 1995
    --------------------------
    Patching file ./src/msql/net.c using Plan A...
    Hunk #1 succeeded at 66.
    Hunk #2 succeeded at 84.
    done
</pre>
 
    If you have problems applying the patch, you may wish to consider
    using some or all of the following arguments to patch: <p>
 
    <ul>
	<li><b>-b</b> <p>
 
	forces <tt>patch</tt> to create a backup of the original file.
	By default it adds a <i>.orig</i> suffix to the original
	filename. <p>
 
	<li><b>-N</b> <p>
 
	forces <tt>patch</tt> to ignore patches that it thinks are
	reversed or already applied. <p>
 
	<li><b>-p0</b> (<i>That's a &quot;zero&quot;, NOT an
	&quot;oh&quot;</i>).<p>
 
	forces <tt>patch</tt> to honour the full pathname of the files
	listed in the patch file.
 
    </ul>
    <p>
 
 
    <li> Recompile mSQL. If you have <a href="#other-apps">third party
    applications</a> that use the mSQL library, you'll probably have to
    <a href="#recompile">recompile</a> them as well.
 
</ol>
 
<hr>
 
<h3><a name="online-doco">Is online documentation available?</a></h3>
 
A set of WWW pages for mSQL, <a href="#w3-msql">W3-mSQL</a>,
<a href="#esl">ESL and Minerva</a> are now available
from <p>
 
<pre>
    <a href="http://Hughes.com.au">http://Hughes.com.au</a>
</pre>
 
Jeff Rowe <a href="mailto:beowulf@cscsun4.larc.nasa.gov">
&lt;beowulf@cscsun4.larc.nasa.gov&gt;</a> has converted the mSQL
documentation to HTML format. It is available from: <p>
 
<pre>
    <a href="http://cscsun1.larc.nasa.gov/~beowulf/msql/msql.html">http://cscsun1.larc.nasa.gov/~beowulf/msql/msql.html</a>
</pre>
 
Tim Finin <a href="mailto:finin@umbc.edu"> &lt;finin@umbc.edu&gt;</a>
has created an online mSQL tutorial. It is available from: <p>
 
<pre>
    <a href="http://www.cs.umbc.edu/~finin/461/msql">http://www.cs.umbc.edu/~finin/461/msql</a>
</pre>
 
Mike Carpenter <a href="mailto:mikec@emi.net">
&lt;mikec@emi.net&gt;</a> has begun to collate the available mSQL
documentation into a single site. He hopes to have a wide range of
resources available for value added mSQL products such as the Perl and
Java interfaces. His work can be found at: <p>
 
<pre>
    <a href="http://www.emi.net/~mikec/tech/msql.html">http://www.emi.net/~mikec/tech/msql.html</a>
</pre>
 
<hr>
 
<h3><a name="mailing-list">Is there a mailing list for mSQL?</a></h3>
 
Yes. <p>
 
To subscribe to the list, send the word &quot;<tt>subscribe</tt>&quot;
in the body of your message to
<a href="mailto:msql-list-request@bunyip.com">
msql-list-request@bunyip.com.</a> <p>
 
To unsubscribe from the list send the word
&quot;<tt>unsubscribe</tt>&quot; in the body of your message to
<a href="mailto:msql-list-request@bunyip.com">
msql-list-request@bunyip.com.</a> <p>
 
Postings should be addressed to <a href="mailto:msql-list@bunyip.com">
msql-list@bunyip.com.</a> <p>
 
<a name="list-archives">Archives</a> of the mailing list are available
via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.bunyip.com/pub/mailing-lists/msql-list.archive/">ftp://ftp.bunyip.com/pub/mailing-lists/msql-list.archive/</a>
</pre>
 
 
Each month's archive is stored in a file: <p>
 
<pre>
    msql-list.archive.YYMM
</pre>
 
where <tt>YYMM</tt> represents the year and month. So the archive for
October 1995 would be in the file: <p>
 
<pre>
    msql-list.archive.9510
</pre>
 
These files are also available from the
<a href="http://www.math.psu.edu/barr/majordomo-faq.html">majordomo</a>
mailing list server at <a href="http://www.bunyip.com">bunyip.com</a>.
To receive a list of the archive files available as well as the
majordomo help file send a message to
<a href="mailto:majordomo@bunyip.com">majordomo@bunyip.com</a> with the
text: <p>
 
<pre>
    index msql-list
    help
    END
</pre>
 
in the body of the message. <p>
 
To reach a human for help regarding the mailing list send a note to:
<p>
 
<pre>
    <a href="mailto:owner-msql-list@bunyip.com">owner-msql-list@bunyip.com</a>
</pre>
 
or
 
<pre>
    <a href="mailto:listmaster@bunyip.com">listmaster@bunyip.com</a>
</pre>
 
The mailing list discusses issues that arise from the use of mSQL and
W3-mSQL (both products developed by David Hughes). Often discussions on
contributed software arise but it is probably best to take these
discussions off line and summarise solutions back to the list. <p>
 
Traffic on the list is fairly high. There were approximately 1100
postings between April 1 1996 and May 31 1996 which gives an average of
around 550 messages per month. (If you think this is high, try
subscribing to the firewalls list - this has an average of around 1000
postings per month!) <p>
 
Turn around times for postings can sometimes be a little slow. It is
not unusual for messages sent from Australia to take a few hours to
appear on the list. List subscribers from other countries have also
reported similar turn around times. Please be patient. <p>
 
<hr>
 
<h3><a name="mailing-list-digest">Is there a mailing list digest for
mSQL?</a></h3>
 
Yes.
 
To subscribe to the list digest, send the words &quot;<tt>subscribe
msql-list-digest</tt>&quot; in the body of your message to
<a href="mailto:majordomo@bunyip.com">majordomo@bunyip.com</a>. <p>
 
To unsubscribe from the list digest send the words &quot;<tt>unsubscribe
msql-list-digest</tt>&quot; in the body of your message to
<a href="mailto:majordomo@bunyip.com">majordomo@bunyip.com</a>. <p>
 
<hr>
 
<h3><a name="html-mail">Are there any HTML archives of the mSQL mailing
list?</a></h3>
 
A number of sites have set up a web based archive of the mSQL mailing
list (among others). Their locations and features are summarised
below: <p>
 
<center>
<table border=2 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<th>
	    Archive Location
	</th>
 
	<th>
	    Features
	</th>
    </tr>
 
    <tr>
	<td>
	    <a href="http://tacyon.spectrum.com.au/mail/msql">
	    http://tacyon.spectrum.com.au/mail/msql</a> <br>
	    (Thanks to Matt Perkins
	    <a href="mailto:matt@spectrum.com.au">
	    &lt;matt@spectrum.com.au&gt;</a>).
	</td>
 
	<td>
	    Threaded archive, list archive, old list archives.
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="http://cure.medinfo.org/lists/programming/index.html">
	    http://cure.medinfo.org/lists/programming/index.html</a> <br>
	    (Thanks to the folks at Medinfo <a href="http://cure.medinfo.org">
	    http://cure.medinfo.org</a>).
	</td>
 
	<td>
	    Threaded archive, browse current month, search list
	    archive. <b>Currently Unavailable.</b>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="http://www.nexial.nl/cgi-bin/msql">
	    http://www.nexial.nl/cgi-bin/msql</a> <br>
	    (Thanks to Kim Hendrikse <a href="mailto:kim@nexial.nl">
	    &lt;kim@nexial.nl&gt;</a>).
	</td>
 
	<td>
	    Fuzzy searchable archive.
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="http://www.findmail.com/listsaver/msql-list">
	    http://www.findmail.com/listsaver/msql-list</a> <br>
	    (Thanks to FindMail Communications
	    <a href="http://www.findmail.com">
	    http://www.findmail.com</a>).
	</td>
 
	<td>
	    Listed by subject thread, date or author. Archives from
	    October 1995.
	</td>
    </tr>
 
</table>
</center>
<p>
 
<hr>
 
<h3><a name="news-group">Is there a news group for mSQL?</a></h3>
 
No. <p>
 
This question comes up about every three months  or so - usually from
new subscribers to the <a href="#mailing-list">mSQL mailing list</a>
who haven't seen the previous threads. <p>
 
There seem to be two distinct groups of people involved in this
discussion: <p>
 
<ol>
    <li> Those who prefer the mailing list. <p>
 
	<ul>
	    <li> News turn around times can be horrendously slow. It
	    is not uncommon for messages to appear in Australia up to
	    seven (7) days after they were originally posted.
	    Considering that mSQL is developed and supported in
	    Australia, a news group would pose an unacceptable delay
	    in providing support for mSQL. <p>
 
	    <li> The mailing list is well established and even if a
	    news group were created, the mailing list would still
	    live on. Creation of a news group would mean that users
	    would need to track both forums in order to keep abreast
	    of current developments in mSQL. <p>
 
	    <li> Newsgroup messages expire - if you happen to be away
	    for an extended period you run the risk of losing valuable
	    information that may have expired while you were away. <p>
 
	    <li> News groups are often subject to &quot;spamming&quot;
	    whereas mailing lists are less vulnerable. <p>
 
	    <li> Signal to noise ratios on news groups tend to be
	    lower than mailing lists (even when instances of
	    &quot;spamming&quot; are discounted). <p>
 
	    <li> Not everyone who subscribes to the mSQL mailing list
	    has a news feed. Establishment of a news group would
	    immediately isolate those mSQL users from any discussions
	    carried out within the news group. <p>
 
            <li> Not everyone has access to a full news feed. Some
            people have difficulty in obtaining permission to receive
            new news groups. Establishment of a news group would
            immediately isolate those mSQL users from any discussions
            carried out within the news group.
	</ul>
	<p>
 
    <li> Those who would prefer a news group. <p>
 
	<ul>
	    <li> A news group allows mSQL related messages to be
	    isolated from other messages. This can be resolved by
	    using mail filtering or threaded mail readers. Examples of
	    mail filtering include <tt>procmail</tt> and <tt>filter</tt>
	    (which is distributed with the Elm mail reader). Release
	    2.<i>x</i> and above of the Netscape Navigator includes a
	    threaded mail reader. <p>
 
	    <li> The volume of messages on the mSQL list is
	    increasing, making it difficult to keep track of threads.
	    This can be alleviated using the methods described above.
	    <p>
 
            <li> It is often more cost effective for some users to
            receive blocks of news rather than individual mail messages.
            Similar behaviour can be achieved by subscribing to the
	    <a href="#mailing-list-digest">mSQL mailing list digest</a>.
	</ul>
</ol>
 
While opinion is divided among these camps, the dominant force at
present is for the mailing list to continue without the establishment
of a global newsgroup. However, this does not preclude you from
setting up a local news/mail gateway such that incoming mail items are
sent to a local newsgroup at your site and outgoing news postings are
sent to the mailing list. Consult your <tt>nntp</tt> documentation for
more details. <p>
 
Nor does it preclude you from establishing a global news group. There
are well defined channels established within the Usenet community for
the creation of new news groups. <p>
 
One publicly available mail/news gateway is available via anonymous
ftp from:
 
<pre>
    <a href="ftp://ftp.vix.com/pub/inn/contrib/newsgate.tar.Z">ftp://ftp.vix.com/pub/inn/contrib/newsgate.tar.Z</a>
    (443025 bytes)
</pre>
 
It includes the programs <tt>mail2news</tt> and <tt>news2mail</tt>. <p>
 
<hr>
 
<h3><a name="posting">How do I post a question to the</a>
<a href="#mailing-list">mSQL mailing list?</a></h3>
 
Before you post a question to the mSQL mailing list, read the manual
that came with the mSQL distribution as well as the rest of this FAQ.
Also, ask yourself the following question: <p>
 
<pre>
    Is there any way I can find the answer to this question myself?
</pre>
 
If you can figure out a way to simply find the answer, then it will
probably be quicker than asking the list. If you think your answer
would be helpful to others then post a summary to the mailing list.
<p>
 
Postings should be addressed to
<a href="mailto:msql-list@bunyip.com"> msql-list@bunyip.com.</a> <p>
 
<b><font size="+1">IF YOU POST A QUESTION TO THE LIST ASKING FOR HELP, YOU
MUST INCLUDE THE FOLLOWING INFORMATION! <p>
 
<ol>
    <li> The operating system you are using - for example Solaris 2.4 or
    Linux 1.3.1. On Unix like systems, this can usually be obtained by
    running the command: <p>
 
<pre>
    uname -a
</pre>
 
    <li> The processor type - for example sparc, Intel 386, HP PA-RISC,
    mc68020. Often the same operating system may run on different
    processors. <p>
 
    <li> The output of the command: <p>
 
<pre>
    msqladmin version
</pre>
 
    <li> Which, if any, unofficial patches you have applied to the
    version of mSQL you're using. <p>
 
    <li> The name and version details of any <a href="#other-apps">third
    party application</a> you are using with mSQL.
 
</ol>
 
Failure to include these details makes it almost impossible to pinpoint
the cause of your problem.</font></b> <p>
 
<hr>
 
<h3><a name="current-bug">I've found a bug. How can I tell if it's been fixed
yet?</a></h3>
 
The status of the development is reflected in the HISTORY file. An
on-line copy of the HISTORY file is kept on the Hughes Technology web
pages. This file is updated as each modification is made to the
sources. Any bugs that have been fixed in the code since it was last
release will be mentioned in the on-line release history information.
You can access this file on the web as: <p>
 
<pre>
    <a href="http://Hughes.com.au/product/msql/history.htm">http://Hughes.com.au/product/msql/history.htm</a>
    </pre>
 
<hr>
 
<h3><a name="reporting-bugs">How do I report a bug in mSQL?</a></h3>
 
The first thing to remember is that you should NOT mail David Hughes
(aka Bambi) directly with your questions - this just makes the process
slower. Any questions you have should be sent to the
<a href="#mailing-list">mSQL mailing list</a>. <p>
 
If you want to report a bug, send a report to the mSQL bug reporting
address at
<a href="mailto:msql-bugs@hughes.com.au">msql-bugs@hughes.com.au.</a>
You may also wish to copy your report to the
<a href="#mailing-list">mSQL mailing list</a> at
<a href="mailto:msql-list@bunyip.com">msql-list@bunyip.com.</a> <p>
 
When making your bug report, please include the following information:
<p>
 
<ul>
    <li> the operating system and version - for example Solaris 2.4 or
    HP-UX 9.05
 
    <li> the processor you are using - for example sparc or Intel 486
 
    <li> include the output of &quot;<tt>msqladmin version</tt>&quot;
 
    <li> which, if any, <a href="#recommended-patches">unofficial
    patches</a> you have applied to the version of mSQL you're using
 
    <li> include a dump of a test database using <tt>msqldump</tt>
 
    <li> include an mSQL script that will show the bug
</ul>
 
You should also ensure that you are running the
<a href="#current-version">latest</a> (stable) version of mSQL before
posting a bug report as David fixes bugs in every version. <p>
 
<hr>
 
<h3><a name="contrib-patches">How do I contribute a patch to
mSQL?</a></h3>
 
Following this procedure will standardise the manner in which unofficial
mSQL patches are distributed to the mSQL user community. <p>
 
<ol>
    <li> Ensure that you are patching the
    <a href="#current-version">latest</a> version of mSQL. <p>
 
    <li> Ensure that you have applied any necessary
    <a href="#recommended-patches"> unofficial recommended patches</a> to
    the latest version. <p>
 
    <li> Make your changes to the &quot;<tt>src</tt>&quot; hierarchy of
    mSQL. For example, if you need to patch the file: <p>
 
<pre>
    msqld.c
</pre>
 
    Save the original file as follows: <p>
 
<pre>
    cp ./src/msql/msqld.c ./src/msql/msqld.c.orig
</pre>
 
    Make your changes to the file: <p>
 
<pre>
    ./src/msql/msqld.c
</pre>
 
    <li> Test, double test and triple test your patch to make sure it
    <b>REALLY</b> fixes a problem. <p>
 
    <li> Generate a difference listing suitable for use by Larry Wall's
    <tt>patch</tt> program:
 
<pre>
    diff -c ./src/msql/msqld.c.orig ./src/msql/msqld.c
</pre>
 
    <li> Post this difference listing and an explanation of the patch to
    the <a href="#mailing-list">mSQL mailing list</a>.
 
</ol>
 
If at all possible, try to avoid posting 'monster' patches. If your
patch modifies a number of separate aspects of mSQL, split each of these
into different patches. This makes it easier for people to apply
feature specific patches to mSQL while avoiding unwanted feature
patches. <p>
 
<hr>
 
<h3><a name="contrib-code">How do I contribute code to the mSQL
community?</a></h3>
 
Contributed code can be uploaded to the ftp server at
<a href="http://bond.edu.au">Bond University</a> Australia. Place your
code in the directory: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Incoming">ftp://bond.edu.au/pub/Minerva/msql/Incoming</a>
</pre>
 
then notify David at
<a href="mailto:bambi@hughes.com.au">&lt;bambi@hughes.com.au&gt;</a>.
He will move your contribution to the mSQL
<a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib">contributions</a>
directory: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib">ftp://bond.edu.au/pub/Minerva/msql/Contrib</a>
</pre>
 
<b>Note:</b> The directory permissions for
<a href="ftp://bond.edu.au/pub/Minerva/msql/Incoming">
<tt>ftp://bond.edu.au/pub/Minerva/msql/Incoming</tt></a> prohibit the
viewing of any files contained therein. This is a security feature
implemented by the system administration staff at
<a href="http://bond.edu.au">Bond University</a> to eliminate recent
attempts to use the ftp server as a relay for the unauthorised transfer
of commercial software. If you upload any software to this directory,
you must notify
<a href="mailto:bambi@hughes.com.au">&lt;bambi@hughes.com.au&gt;</a> to
have it moved to <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib">
<tt>ftp://bond.edu.au/pub/Minerva/msql/Contrib</tt></a>. <p>
 
You may like to discuss your proposed code with others on the
<a href="#mailing-list">mSQL mailing list</a>. The subscribers to this
list may be able to help you with improvements or modifications to your
code or advise you of work already available in your area. <p>
 
<hr>
 
<h3><a name="code-examples">Where can I find examples of code that
interface with mSQL?</a></h3>
 
If you're writing code in the C programming language, examine the mSQL
distribution itself. All the auxiliary programs use the mSQL
<a href="#api">API</a> in a variety of ways. Pay particular attention
to: <p>
 
<ul>
    <li> <a href="#support-progs"><tt>./src/msql/msql.c</tt></a>
    <li> <a href="#support-progs"><tt>./src/msql/relshow.c</tt></a>
    <li> <a href="#support-progs"><tt>./src/msql/msqldump.c</tt></a>
    <li> <a href="#support-progs"><tt>./src/msql/msqladmin.c</tt></a>
</ul>
 
If you're writing code in other languages, have a look through the
distribution of the mSQL language extension itself for examples.
Another good place to look is the <a href="#list-archives">mSQL mailing
list archives</a>. <p>
 
<hr>
 
<h3><a name="porting">Are there any non Unix ports of mSQL?</a></h3>
 
Both mSQL version 1.<i>x</i> and 2.<i>x</i> have been ported to a few
other operating systems - namely MS Windows, MS Windows 95, MS Windows
NT and OS/2. For more details on these ports and any other ports,
please see the <a href="#other-apps">&quot;Contributed Code and Third
Party Applications&quot;</a> section of this FAQ. <p>
 
<hr>
<hr width="75%">
 
<!-- End of Part 1 -->
 
<center>
<h2><a name="features-and-limits">Features and Limitations</a></h2>
</center>
 
<h3><a name="support-progs">What support programs are bundled with
mSQL?</a></h3>
 
mSQL is bundled with the following programs: <p>
 
<center>
<table border=2 cellpadding=3 width="50%" bgcolor="#cccccc">
    <tr>
	<th colspan=2>Support Programs bundled with mSQL version
	1.<i>x</i></th>
    </tr>
 
    <tr>
	<td align=center><b>msqld</b></td>
	<td>the mSQL database server.</td>
    </tr>
 
    <tr>
	<td align=center><b>msqladmin</b></td>
	<td>handles administrative details such as creation and
	deletion of databases, server shutdown etc.</td>
    </tr>
 
    <tr>
	<td align=center><b>msql</b></td>
	<td>the mSQL monitor. It provides a user interface for making
	SQL queries on databases.</td>
    </tr>
 
    <tr>
	<td align=center><b>msqldump</b></td>
	<td>Dumps a database in ASCII format. The dump takes the form
	of SQL commands so that the entire database can be re-created
	from the dump file.</td>
    </tr>
 
    <tr>
        <td align=center><b>relshow</b></td>
	<td>The mSQL schema viewer.  Shows <a href="#schema">table
	details</a> for a given database.</td>
    </tr>
</table>
</center>
<p>
 
<center>
<table border=2 cellpadding=3 width="50%" bgcolor="#cccccc">
    <tr>
	<th colspan=2>Support Programs bundled with mSQL version
	2.<i>x</i></th>
    </tr>
 
    <tr>
	<td align=center><b>msql2d</b></td>
	<td>the mSQL database server.</td>
    </tr>
 
    <tr>
	<td align=center><b>msqladmin</b></td>
	<td>handles administrative details such as creation
	and deletion of databases, server shutdown etc.</td>
    </tr>
 
    <tr>
	<td align=center><b>msql</b></td>
	<td>the mSQL monitor. It provides a user interface
	for making SQL queries on databases.</td>
    </tr>
 
    <tr>
	<td align=center><b>msqldump</b></td>
	<td>Dumps a database in ASCII format. The dump
	takes the form of SQL commands so that the entire database can
	be re-created from the dump file.</td>
    </tr>
 
    <tr>
	<td align=center><b>msqlexport</b></td>
	<td>Dumps a database in ASCII format. The dump
	takes the form of user defined character delimited fields
	suitable for importing into other vendor's databases.</td>
    </tr>
 
    <tr>
	<td align=center><b>msqlimport</b></td>
	<td>Loads a flat file in ASCII format into an mSQL database.</td>
    </tr>
 
    <tr>
        <td align=center><b>relshow</b></td>
	<td>The mSQL schema viewer. Shows
	<a href="#schema">table details</a> for a given database.</td>
    </tr>
 
    <tr>
	<td align=center><b>w3-msql</b></td>
	<td>CGI program that allows users to embed mSQL
	statements in their HTML documents. </td>
    </tr>
 
    <tr>
	<td align=center><b>w3-auth</b></td>
	<td>Access control and authentication module for
	use with <tt>w3-msql</tt></td>
    </tr>
 
    <tr>
	<td align=center><b>lite</b></td>
	<td>A stand alone mSQL scripting language. </td>
    </tr>
</table>
</center>
<p>
 
For more details on these programs see the documentation that comes
with mSQL. <p>
 
<hr>
 
<h3><a name="sql-commands">What SQL commands are supported?</a></h3>
 
mSQL version 1.<i>x</i> supports the following SQL commands: <p>
 
<center>
<table border=2 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<th>
	    mSQL version 1.<i>x</i>
	</th>
    </tr>
 
    <tr>
	<td>
<pre>
CREATE TABLE <i>table_name</i> (
    <i>col_name</i> col_type [ not null | primary key ]
    [, <i>col_name</i> col_type [ not null | primary key ] ]**
)
</pre>
	</td>
    </tr>
 
    <tr>
	<td>
<pre>
DROP TABLE <i>table_name</i>
</pre>
	</td>
    </tr>
 
    <tr>
	<td>
<pre>
INSERT INTO <i>table_name</i> [ ( <i>column</i> [ , <i>column</i> ]** ) ]
    VALUES (<i>value</i> [, <i>value</i>]** )
</pre>
	</td>
    </tr>
 
    <tr>
	<td>
<pre>
DELETE FROM <i>table_name</i>
    WHERE <i>column</i> OPERATOR <i>value</i>
    [ AND | OR <i>column</i> OPERATOR <i>value</i> ]**
 
	OPERATOR can be &lt;, &gt;, =, &lt;=, &gt;=, &lt;&gt;, or LIKE
</pre>
	</td>
    </tr>
 
    <tr>
	<td>
<pre>
SELECT [ DISTINCT ] [<i>table</i>.]<i>column</i> [ , [<i>table</i>.]<i>column</i> ]**
    FROM <i>table</i> [ = <i>alias</i>] [ , <i>table</i> [ = <i>alias</i>] ]**
    [ WHERE [<i>table</i>.]<i>column</i> OPERATOR VALUE
	[ AND | OR [<i>table</i>.]<i>column</i> OPERATOR VALUE]** ]
    [ ORDER BY [<i>table</i>.]<i>column</i> [DESC] [, [<i>table</i>.]<i>column</i> [DESC] ] [LIMIT <i>n</i>]
 
	OPERATOR can be &lt;, &gt;, =, &lt;=, &gt;=, &lt;&gt;, or LIKE
	VALUE can be a literal value or a column name
</pre>
	</td>
    </tr>
 
    <tr>
	<td>
<pre>
UPDATE <i>table_name</i> SET <i>column</i>=<i>value</i> [ , <i>column</i>=<i>value</i> ]**
    WHERE <i>column</i> OPERATOR <i>value</i>
    [ AND | OR <i>column</i> OPERATOR <i>value</i> ]**
 
	OPERATOR can be &lt;, &gt;, =, &lt;=, &gt;=, &lt;&gt;, or LIKE
</pre>
	</td>
    </tr>
</table>
</center>
<p>
 
For more details see the documentation that comes with mSQL version
1.<i>x</i>. <p>
 
mSQL version 2.<i>x</i> supports the following SQL commands: <p>
 
<center>
<table border=2 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<th>
	    mSQL version 2.<i>x</i>
	</th>
    </tr>
 
    <tr>
	<td>
<pre>
CREATE TABLE <i>table_name</i> (
    <i>col_name</i> col_type [ not null ]
    [, <i>col_name</i> col_type [ not null ] ]**
)
</pre>
	</td>
    </tr>
 
 
    <tr>
	<td>
<pre>
CREATE [ UNIQUE ] INDEX <i>index_name</i> ON <i>table_name</i>  (
    <i>col_name</i> [, <i>col_name</i> ]**
)
</pre>
	</td>
    </tr>
 
 
    <tr>
	<td>
<pre>
CREATE SEQUENCE ON <i>table_name</i>
    [ STEP <i>step_val</i> ] [ VALUE <i>initial_val</i> ]
</pre>
	</td>
    </tr>
 
    <tr>
	<td>
<pre>
DROP TABLE <i>table_name</i>
</pre>
	</td>
    </tr>
 
 
    <tr>
	<td>
<pre>
DROP INDEX <i>index_name</i> FROM <i>table_name</i>
</pre>
	</td>
    </tr>
 
 
    <tr>
	<td>
<pre>
DROP SEQUENCE FROM <i>table_name</i>
</pre>
	</td>
    </tr>
 
    <tr>
	<td>
<pre>
INSERT INTO <i>table_name</i> [ ( <i>column</i> [ , <i>column</i> ]** ) ]
    VALUES (<i>value</i> [, <i>value</i>]** )
</pre>
	</td>
    </tr>
 
    <tr>
	<td>
<pre>
DELETE FROM <i>table_name</i>
    WHERE <i>column</i> OPERATOR <i>value</i>
    [ AND | OR <i>column</i> OPERATOR <i>value</i> ]**
 
	OPERATOR can be &lt;, &gt;, =, &lt;=, &gt;=, &lt;&gt;, LIKE, RLIKE or CLIKE
</pre>
	</td>
    </tr>
 
    <tr>
	<td>
<pre>
SELECT [ DISTINCT ] [<i>table</i>.]<i>column</i> [ , [<i>table</i>.]<i>column</i> ]**
    FROM <i>table</i> [ = <i>alias</i>] [ , <i>table</i> [ = <i>alias</i>] ]**
    [ WHERE [<i>table</i>.]<i>column</i> OPERATOR VALUE
	[ AND | OR [<i>table</i>.]<i>column</i> OPERATOR VALUE]** ]
    [ ORDER BY [<i>table</i>.]<i>column</i> [DESC] [, [<i>table</i>.]<i>column</i> [DESC] ] [LIMIT <i>n</i>]
 
	OPERATOR can be &lt;, &gt;, =, &lt;=, &gt;=, &lt;&gt;, LIKE, RLIKE or CLIKE
	VALUE can be a literal value or a column name
	WHERE clauses may contain parentheses () to nest conditions
</pre>
	</td>
    </tr>
 
    <tr>
	<td>
<pre>
UPDATE <i>table_name</i> SET <i>column</i>=<i>value</i> [ , <i>column</i>=<i>value</i> ]**
    WHERE <i>column</i> OPERATOR <i>value</i>
    [ AND | OR <i>column</i> OPERATOR <i>value</i> ]**
 
	OPERATOR can be &lt;, &gt;, =, &lt;=, &gt;=, &lt;&gt;, LIKE, RLIKE or CLIKE
</pre>
	</td>
    </tr>
</table>
</center>
<p>
 
For more details see the documentation that comes with mSQL version
2.<i>x</i>. <p>
 
<hr>
 
<h3><a name="keys">Does mSQL support keys?</a></h3>
 
Yes. <p>
 
Under mSQL version 1.<i>x</i> any single field of a table can be
designated as the <tt>PRIMARY KEY</tt>. These keys are, by definition,
unique. mSQL version 1.<i>x</i> does not support multiple keys within
a table.  <p>
 
The creation of a <tt>PRIMARY KEY</tt> is made during the table
declaration, for example: <p>
 
<pre>
    CREATE TABLE employee (
	employee_number		INT PRIMARY KEY,
	department		CHAR(20),
	first_name		CHAR(20) NON NULL,
	last_name		CHAR(20) NON NULL
    )
 
    CREATE TABLE department (
	department_name		CHAR(20) PRIMARY KEY
    )
</pre>
 
For more details see the documentation that comes with mSQL version
1.<i>x</i>. <p>
 
mSQL version 2.<i>x</i> has disposed of the <tt>PRIMARY KEY</tt>
in favour of an indexing mechanism. <p>
 
An <tt>INDEX</tt> may be created for any field of a table at any time,
for example: <p>
 
<pre>
    CREATE UNIQUE INDEX idx1 ON employee(employee_number)
    CREATE INDEX idx2 ON employee(first_name, last_name)
</pre>
 
This examples shows that the first <tt>INDEX</tt>, <i>idx1</i>, is
unique. This is identical to the <tt>PRIMARY KEY</tt> concept used by
mSQL version 1.<i>x</i>. The second <tt>INDEX</tt>, <i>idx2</i>, need
not be unique. <i>idx2</i> also spans two fields within a table which
will improve performance whenever a search is made based on these two
fields. <p>
 
These indices will be used automatically whenever a query is sent to
the database engine that uses those fields in its <tt>WHERE</tt>
clause. The user is not required to specify any special values in the
query to ensure the indices are used to increase performance. <p>
 
<hr>
 
<h3><a name="datatypes">What datatypes are supported?</a></h3>
 
mSQL version 1.<i>x</i> supports the following datatypes: <p>
 
<pre>
    int		(4 bytes)	-2147483646 &lt;= i &lt;= 2147483647
    real	(8 bytes)	4.94E-324 &lt;= x &lt;= 1.79E+308, -1.79E+308 &lt;= x &lt;= -4.94E-324
    char
</pre>
 
The internal storage for types <tt>int</tt> and <tt>real</tt> is held at
4 bytes and 8 bytes respectively, regardless of the system architecture
you're using. So even on 64 bit Crays a <tt>real</tt> will be 8 bytes.
<p>
 
There is NO fixed limit on the size of a character field. Provided you
declare it correctly when the table is defined, mSQL will handle
fields of at least several thousand characters without problems. You
may run into difficulty actually using fields of this size in whatever
language you choose to interface to the mSQL database engine. <p>
 
<b>Note:</b> If you declare a <tt>char</tt> field with a large size,
each and every row in the table will allocate sufficient storage space
for the <tt>char</tt> field - whether the full size of the field is
used or not. This may lead to very large database tables indeed. <p>
 
An example of each mSQL version 1.<i>x</i> datatype within a table
declaration follows: <p>
 
<pre>
    CREATE TABLE <i>table_name</i> (
	    <i>field_name_1</i>	INT PRIMARY KEY,
	    <i>field_name_2</i>	REAL,
	    <i>field_name_3</i>	CHAR(10) NOT NULL,
	    <i>field_name_4</i>	CHAR(80),
	    <i>field_name_5</i>	INT	NOT NULL,
	    <i>field_name_6</i>	REAL
	)
</pre>
 
For more details see the documentation that comes with mSQL version
1.<i>x</i>. <p>
 
mSQL version 2.<i>x</i> supports the following datatypes: <p>
 
<pre>
    int		(4 bytes)	-2147483646 &lt;= i &lt;= 2147483647
    uint	(4 bytes)	0 &lt;= i &lt;= 4294967295
    real	(8 bytes)	4.94E-324 &lt;= x &lt;= 1.79E+308, -1.79E+308 &lt;= x &lt;= -4.94E-324
    char
    text
    date			dd-mon-yy, for example 12-May-97
    time
    money
</pre>
 
The internal storage for types <tt>int/uint</tt> and <tt>real</tt> is
held at 4 bytes and 8 bytes respectively, regardless of the system
architecture you're using. So even on 64 bit Crays a <tt>real</tt> will
be 8 bytes.  <p>
 
There is NO fixed limit on the size of a character field. Provided you
declare it correctly when the table is defined, mSQL will handle
fields of at least several thousand characters without problems. You
may run into difficulty actually using fields of this size in whatever
language you choose to interface to the mSQL database engine. <p>
 
<b>Note:</b> If you declare a <tt>char</tt> field with a large size,
each and every row in the table will allocate sufficient storage space
for the <tt>char</tt> field - whether the full size of the field is
used or not. This may lead to very large database tables indeed. <p>
 
<tt>text</tt> (or other 8 bit data) fields are defined with an expected
average length. Values longer than the specified length will be split
between the data table and external overflow buffers. <tt>text</tt>
fields are slower to access than <tt>char</tt> fields and cannot be
used in an <tt>index</tt> or in <tt>LIKE</tt> tests. <p>
 
An example of each mSQL version 2.<i>x</i> datatype within a table
declaration follows: <p>
 
<pre>
    CREATE TABLE <i>table_name</i> (
	    <i>field_name_1</i>		INT,
	    <i>field_name_2</i>		REAL,
	    <i>field_name_3</i>		CHAR(10)	NOT NULL,
	    <i>field_name_4</i>		CHAR(80),
	    <i>field_name_5</i>		INT		NOT NULL,
	    <i>field_name_6</i>		REAL,
	    <i>field_name_7</i>		TEXT(25),
	    <i>field_name_8</i>		DATE,
	    <i>field_name_9</i>		UINT,
	    <i>field_name_10</i>		TIME,
	    <i>field_name_11</i>		MONEY
	)
</pre>
 
<b>Note:</b> The storage of real numbers is highly machine dependent.
If you store the number <tt>10.03</tt> don't be surprised if it is
actually stored as either <tt>10.03</tt> or <tt>10.03000000001</tt> or
<tt>10.02999999999</tt>. <p>
 
<b>Note:</b> The upper and lower limits for real numbers are also
highly machine dependent. Both mSQL version 1.<i>x</i> and 2.<i>x</i>
obtain these values from the system include file: <p>
 
<pre>
    /usr/include/limits.h
</pre>
 
If you plan on storing very big numbers (positive or negative numbers
with exponents approaching 308) or very small numbers (positive or
negative numbers with exponents approaching -324) you <b>MUST</b> check
whether your operating system will support them. <p>
 
<b>Note:</b> When dealing with real numbers it might be advisable to
either convert the real number to an integer if possible. For example
if you are dealing with decimal currency, it may be preferable to work
in units of cents rather than dollars - <b>note</b> that mSQL
2.<i>x</i> now has a <tt>MONEY</tt> type which handles this concept -
or to perform some sort of delta check when retrieving values. The
following logic demonstrates one possible application of this concept:
<p>
 
<pre>
    delta = 0.00001
    select a real number from a table
 
    if abs(number - expected value) &lt; delta
    then
	proceed
    else
	fail
</pre>
 
<hr>
 
<h3><a name="api">What functions does the mSQL API provide?</a></h3>
 
The mSQL API provides the following functions and macros: <p>
 
<center>
<table border=2 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<th colspan=3>mSQL version 1.<i>x</i></th>
    </tr>
 
    <tr>
	<th>Return Type</th>
	<th>Prototype</th>
	<th>Type</th>
    </tr>
 
    <tr>
	<td>int</td>
	<td>msqlConnect(char *host);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>int</td>
	<td>msqlSelectDB(int sock, char *dbname);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>int</td>
	<td>msqlQuery(int sock, char *query);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>m_result *</td>
	<td>msqlStoreResult();</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>void</td>
	<td>msqlFreeResult(m_result *result);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>m_row</td>
	<td>msqlFetchRow(m_result *result);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>void</td>
	<td>msqlDataSeek(m_result *result, int pos);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>int</td>
	<td>msqlNumRows(m_result *result);</td>
	<td align=center>MACRO</td>
    </tr>
 
    <tr>
	<td>m_field *</td>
	<td>msqlFetchField(m_result *result);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>void</td>
	<td>msqlFieldSeek(m_result *result, int pos);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>int</td>
	<td>msqlNumFields(m_result *result);</td>
	<td align=center>MACRO</td>
    </tr>
 
    <tr>
	<td>m_result *</td>
	<td>msqlListDBs(int sock);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>m_result *</td>
	<td>msqlListTables(int sock);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>m_result *</td>
	<td>msqlListFields(int sock, char *tableName);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>void</td>
	<td>msqlClose(int sock);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
</table>
</center>
<p>
 
For more details see the documentation that comes with mSQL version
1.<i>x</i>. <p>
 
<b>Note:</b> The PostScript documentation that comes with mSQL version
1.<i>x</i> lists the function <tt>msqlClose()</tt> as type
<tt>int</tt>. This is incorrect. It is actually type <tt>void</tt>.
<p>
 
<center>
<table border=2 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<th colspan=3>mSQL version 2.<i>x</i></th>
    </tr>
 
    <tr>
	<th>Return Type</th>
	<th>Prototype</th>
	<th>Type</th>
    </tr>
 
    <tr>
	<td>int</td>
	<td>msqlConnect(char *host);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>int</td>
	<td>msqlSelectDB(int sock, char *dbname);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>int</td>
	<td>msqlQuery(int sock, char *query);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>m_result *</td>
	<td>msqlStoreResult();</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>void</td>
	<td>msqlFreeResult(m_result *result);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>m_row</td>
	<td>msqlFetchRow(m_result *result);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>void</td>
	<td>msqlDataSeek(m_result *result, int pos);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>int</td>
	<td>msqlNumRows(m_result *result);</td>
	<td align=center>MACRO</td>
    </tr>
 
    <tr>
	<td>m_field *</td>
	<td>msqlFetchField(m_result *result);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>void</td>
	<td>msqlFieldSeek(m_result *result, int pos);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>int</td>
	<td>msqlNumFields(m_result *result);</td>
	<td align=center>MACRO</td>
    </tr>
 
    <tr>
	<td>m_seq *</td>
	<td>msqlGetSequenceInfo(int sock, char *table);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>m_result *</td>
	<td>msqlListDBs(int sock);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>m_result *</td>
	<td>msqlListTables(int sock);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>m_result *</td>
	<td>msqlListFields(int sock, char *tableName);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>m_result *</td>
	<td>msqlListIndex(int sock, char *tableName, char *index);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>void</td>
	<td>msqlClose(int sock);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>char</td>
	<td>msqlUnixTimeToDate(char *date);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>char</td>
	<td>msqlUnixTimeToTime(char *date);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>time_t</td>
	<td>msqlDateToUnixTime(char *date);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
    <tr>
	<td>time_t</td>
	<td>msqlTimeToUnixTime(char *date);</td>
	<td align=center>FUNCTION</td>
    </tr>
 
</table>
</center>
<p>
 
For more details see the documentation that comes with mSQL version
2.<i>x</i>. <p>
 
<b>Note:</b> The HTML documentation that comes with mSQL version
2.<i>x</i> lists the function <tt>msqlClose()</tt> as type
<tt>int</tt>. This is incorrect. It is actually type <tt>void</tt>.
<p>
 
<hr>
 
<h3><a name="views">Are views or virtual tables supported?</a></h3>
 
No. <p>
 
This feature may be included in mSQL version 2. <p>
 
<hr>
 
<h3><a name="table-aliasing">Does mSQL support table aliasing?</a></h3>
 
Yes. <p>
 
As an example, consider this method to find the list of grandparents
from a child/parent tuple: <p>
 
<pre>
    SELECT t1.parent, t2.child from parent_data=t1,parent_data=t2
	WHERE t1.child=t2.parent
</pre>
 
mSQL also supports the SQL standard method of table aliasing which uses
either a space character or the keyword AS instead of an = character.
So the above example can also be written as either: <p>
 
<pre>
    SELECT t1.parent, t2.child from parent_data t1,parent_data t2
	WHERE t1.child=t2.parent
</pre>
 
or
 
<pre>
    SELECT t1.parent, t2.child from parent_data AS t1,parent_data AS t2
	WHERE t1.child=t2.parent
</pre>
 
<hr>
 
<h3><a name="column-constraints">Are column constraints
supported?</a></h3>
 
No. <p>
 
<hr>
 
<h3><a name="stored-procedures">Are stored procedures
supported?</a></h3>
 
No - that's getting beyond the scope of Mini SQL (see Oracle :) <p>
 
<hr>
 
<h3><a name="acl">Are access privileges supported?</a></h3>
 
Yes. <p>
 
mSQL has an access control file which allows the database administrator
to control access on a user and host level. <p>
 
For more details see the documentation that comes with mSQL. <p>
 
mSQL does not support access control from within SQL commands. <p>
 
<hr>
 
<h3><a name="blobs">Does mSQL support BLOBs?</a></h3>
 
BLOBs are <b>B</b>inary <b>L</b>arge <b>OB</b>jects. Typical examples
would be large blocks of text, graphics or audio data. <p>
 
The current release of mSQL has NO direct support for BLOBs. However,
you can always store the path name of a file that points to the BLOB in
one of the fields of your table. Then your application can deal with
the file name appropriately. <p>
 
If you're dealing with large blocks of text, you may also wish to
consider this approach from Pascal Forget
<a href="mailto:pascal@wsc.com">&lt;pascal@wsc.com&gt;</a>: <p>
 
<pre>
    Another possible hack would be to have the &quot;block_text&quot;
    record contain a pointer to a &quot;text_lines&quot; table. This
    table would contain a linked list of text lines like this:
 
	CREATE TABLE text_lines (
	    line_id	int primary key,
	    string	char(80),
	    next_line	int
	)
</pre>
 
Mike Eggleston
<a href="mailto:mikee@sys8.wfc.com">&lt;mikee@sys8.wfc.com&gt;</a>
offers this solution: <p>
 
<pre>
    What I prefer to do in databases where I have text fields containing
    near infinite amounts of text is to define several tables like:
 
	create table prog (
	  id int,
	  name char(40),
	  programmer char(40),
	  ....
	)\p\g
 
	create table descript (
	  id int,
	  line int,
	  descript char(100)
	)\p\g
 
    Then in a program I break up the text as necessary and put one line
    of text in each record of &lt;descript&gt;. When I want it back, by
    program, I [use]
 
	select line, descript from prog, descript
	    where prog.id = descript.id
	    and name = 'foobar' order by line\p\g
</pre>
 
Later versions of mSQL may have support for BLOBs. <p>
 
<hr>
 
<h3><a name="transactions">Are the transaction commands, BEGIN, COMMIT,
and ROLLBACK supported?</a></h3>
 
No, and they will probably never be supported (once again that's beyond
the scope of mSQL).<p>
 
The mSQL server handles requests serially - that is only one user's
request is handled at a time. Therefore there is no danger of a user
reading from a table that is currently being updated by another user.
<p>
 
However, there is the possibility that the same read operations may
straddle another user's write operation so that different data will be
returned from each read. <p>
 
mSQL version 2 will provide client initiated locking. <p>
 
<hr>
 
<h3><a name="table-limits">What are the limits on table and field
names?</a></h3>
 
A table or field name MUST begin with an alphabetic character. The
remainder of the name may be any of the following 63 characters: <p>
 
<ul>
    <li> A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
 
    <li> a b c d e f g h i j k l m n o p q r s t u v w x y z
 
    <li> 0 1 2 3 4 5 6 7 8 9
 
    <li> _ <i>(that's an &quot;underbar&quot; NOT a
    &quot;dash&quot;)</i>
</ul>
 
The maximum name length is set to 20 characters. That's for a table or
a field in a table. This allows for
813621925049196536663393538834956800 possible naming combinations using
the characters listed above. For the mathematically inclined that's:
<p>
 
<pre>
	   19
	 -----
	 \
	  \        i
   52 *   /      63   =   813621925049196536663393538834956800
	 /
	 -----
	 i = 0
</pre>
 
Though in practise, many of these combinations will probably remain
unused. <p>
 
While not recommended, the default maximum name length value of 20 can
be changed by editing the mSQL source code. However, if you change it
AFTER you have already created ANY databases, the old databases will be
unreadable. To avoid this problem, follow this procedure: <p>
 
<ol>
 
    <li> Dump your old databases to ASCII files using
    <a href="#support-progs"><tt>msqldump</tt></a>. <p>
 
    <li> Drop your old databases using
    <a href="#support-progs"><tt>msqladmin</tt></a>. <p>
 
    <li> Shutdown the mSQL database server
    <a href="#support-progs"><tt>msqld</tt></a> using
    <a href="#support-progs"><tt>msqladmin</tt></a>. <p>
 
    <li> Edit the mSQL source and modify
 
<pre>
    ./src/msql/msql_priv.h
</pre>
 
    Change the line reading <p>
 
<pre>
    #define NAME_LEN	19		/* Field/table name length */
</pre>
 
    to suit your needs. Ensure that you use a number that is one less
    than the maximum value you desire. For example, if you wish to have
    a name length of 36 you would change the line to read: <p>
 
<pre>
    #define NAME_LEN	35		/* Field/table name length */
</pre>
 
    <li> Recompile and install the modified mSQL. <p>
 
    <li> Start the new mSQL database server
    <a href="#support-progs"><tt>msqld</tt></a>. <p>
 
    <li> Create new databases using
    <a href="#support-progs"><tt>msqladmin</tt></a>. <p>
 
    <li> Repopulate your databases using
    <a href="#support-progs"><tt>msql</tt></a> and the ASCII dumps from
    step 1.
 
</ol>
 
<hr>
 
<h3><a name="user-limits">What other limits can be modified?</a></h3>
 
The file <p>
 
<pre>
    ./src/msql/msql_priv.h
</pre>
 
contains the definitions of the internal mSQL limits: <p>
 
<pre>
    #define MAX_FIELDS	75		/* Max fields per query */
    #define MAX_CON	24		/* Max connections */
    #define BUF_SIZE	(256*1024)	/* Read buf size if no mmap() */
    #define NAME_LEN	19		/* Field/table name length */
    #define PKT_LEN	(32*1024)	/* Max size of client/server packet */
    #define CACHE_SIZE	8		/* Size of table cache */
</pre>
 
If you want to increase them you can just edit this file and recompile.
Don't change MAX_CON or CACHE_SIZE without understanding why these
limits are set (maximum file descriptors per process etc). <p>
 
Changing any of these parameters will almost certainly make any
existing databases unreadable. To avoid this problem, follow this
procedure: <p>
 
<ol>
 
    <li> Dump your old databases to ASCII files using
	<a href="#support-progs"><tt>msqldump</tt></a>. <p>
 
    <li> Drop your old databases using
	<a href="#support-progs"><tt>msqladmin</tt></a>. <p>
 
    <li> Shutdown the mSQL database server
	<a href="#support-progs"><tt>msqld</tt></a> using
	<a href="#support-progs"><tt>msqladmin</tt></a>. <p>
 
    <li> Edit the mSQL source and modify
 
<pre>
    ./src/msql/msql_priv.h
</pre>
 
    changing the definitions to suit your needs. <p>
 
    <li> Recompile and install the modified mSQL. <p>
 
    <li> Start the new mSQL database server
	<a href="#support-progs"><tt>msqld</tt></a>. <p>
 
    <li> Create new databases using
	<a href="#support-progs"><tt>msqladmin</tt></a>. <p>
 
    <li> Repopulate your databases using
	<a href="#support-progs"><tt>msql</tt></a> and the ASCII dumps
	from step 1.
 
</ol>
 
<hr>
 
<h3><a name="db-size">How much data can mSQL address?</a></h3>
 
mSQL can theoretically address tables with a maximum size of 4
gigabytes. In practise you'll probably run up against operating system
limitations well before this theoretical limit. <p>
 
<hr>
 
<h3><a name="expressions">Are there any limitations in the way mSQL
handles logical expressions?</a></h3>
 
mSQL version 1.<i>x</i> has a very simplistic approach to logical
expressions. <p>
 
Consider the SQL query: <p>
 
<pre>
    SELECT something from somewhere WHERE
	name='jan' or country='italy' and sex='female' or title='ms'
</pre>
 
Under mSQL version 1.<i>x</i>, the parser will scan the condition
from left to right. So in this example the condition reads: <p>
 
<pre>
    ((name='jan' or country='italy') and sex='female') or title='ms'
</pre>
 
mSQL version 1.<i>x</i> does NOT support parentheses in logical
expressions, so there is <b>NO</b> way to change this parsing. <p>
 
Ted Harding <a href="mailto:Ted.Harding@nessie.mcc.ac.uk">
&lt;Ted.Harding@nessie.mcc.ac.uk&gt;</a> provides some solutions for
three component queries. <p>
 
Ted writes: <p>
 
<pre>
    Let's get something clear: ALL 3-component (or equivalent) queries
    can be implemented in mSQL (without parentheses and using the mSQL
    left-to-right evaluation). There are only the following:
 
	    (A and B) and C  =  A and (B and C)  =  A and B and C
	    (A or B) or C    =  A or (B or C)    =  A or B or C
	    (A and B) or C   =  A and B or C
	    A and (B or C)   =  (B or C) and A   =  B or C and A
	    (A or B) and C   =  A or B and C
	    A or (B and C)   =  (B and C) or A   =  B and C or A
 
    Queries like A and B and C and (E or F) are the same as (A and B
    and C) and (E or F) which is the same form as G and (E or F).
 
    The trouble starts with 4-component queries such as (A or B) and (C
    or D) for which there is no one-pass generally correct mSQL
    representation.
</pre>
 
mSQL version 2.<i>x</i> allows parentheses within logical expressions
so this limitation is eliminated. <p>
 
<hr>
 
<h3><a name="return-values">How does mSQL return values?</a></h3>
 
mSQL returns all values as ASCII strings. If, for example, you are
expecting an integer result you may have to do some internal
conversions depending on the language you're using. <p>
 
In C, for example, see the manual pages on <tt>atoi()</tt>. <p>
 
<hr>
 
<h3><a name="return-order">How does SELECT return rows?</a></h3>
 
On Tue, 4 Jul 1995, Karri-Pekka Laakso wrote: <p>
 
<pre>
    Does SELECT return the rows always in order 'first inserted first',
    if there is no ORDER statement given, and the rows are selected
    from one table only, and there has been no DELETEs on that table?
    It seems be so, but is it guaranteed?
</pre>
 
David Hughes replied: <p>
 
<pre>
    This is guaranteed. The only time the rows will be returned in
    another order is if you have deleted a row and it's then filled by
    a later insert.
</pre>
 
<hr>
 
<h3><a name="nested-tables">Can mSQL nest tables?</a></h3>
 
On Tue, 25 Jul 1995, Jerome Stoller wrote: <p>
 
<pre>
    I am new at mSQL, and have a beginner question: Is it possible to
    create a table &quot;normally&quot;, and to have the fields of one
    of the column being[sic] another table?
</pre>
 
David Hughes replied: <p>
 
<pre>
    You can't nest tables in mSQL (don't think you can in ANSI SQL
    either). What you can do is to use a common value as a key to join
    the contents of two tables (eg. a part number or a user ID).
</pre>
 
<hr>
 
<h3><a name="storage-overhead">What storage overheads does mSQL
have?</a></h3>
 
mSQL stores each database in its own directory under the
'<tt>msqldb</tt>' directory of wherever you specified mSQL should be
installed. For example if you specified that mSQL should be installed
in: <p>
 
<pre>
    /usr/local/Minerva/
</pre>
 
then the databases will be created in the directory: <p>
 
<pre>
    /usr/local/Minerva/msqldb/
</pre>
 
<b>Note</b> that this can be overridden by specifying the MSQL_HOME
environment variable when starting
<a href="#support-progs"><tt>msqld</tt></a>. <p>
 
Each table in the database is stored as a number of files: <p>
 
<ul>
    <li><b><i>dbname/table</i>.dat</b> - table data. This file contains
    the data for all the rows in the table. <p>
 
    For each field in a table, mSQL will also store an additional flag
    byte. mSQL also stores an additional flag byte for each row of the
    table. <p>
 
    Consider the following table: <p>
 
<pre>
    CREATE TABLE test (
	f0	char(13),
	f1	int,
	f2	real,
	f3	real,
	f4	real,
	f5	real,
	f6	int
    )
</pre>
 
    Storage space for each row of this table would be: <p>
 
<pre>
    (13 * char) + (2 * int) + (4 * double) + (7 * fields) + (1 * rows)
	=  (13 * 1) + (2 * 4) + (4 * 8) + (7 * 1) + 1
	=  61 bytes
</pre>
 
    So if this table had 1000 records, it would occupy 61000 bytes of
    disk space. (In reality it may occupy slightly more real disk space
    because of the way the underlying file system behaves. This is
    operating system specific and not really an issue to worry about.
    If you do an '<tt>ls -l</tt>' on the file it will show 61000
    bytes). <p>
 
    <li><b><i>dbname/table</i>.key</b> - table keys. This file will
    only exist if the table has a primary key. It contains the key
    pointers into the table data file. <p>
 
    The size of this file will be the size of the key plus one flag
    byte times the number of rows in the table. In the above example,
    if the table was defined as: <p>
 
<pre>
    CREATE TABLE test (
	f0	char(13) primary key,
	f1	int,
	f2	real,
	f3	real,
	f4	real,
	f5	real,
	f6	int
    )
</pre>
 
    and the table had 1000 rows, the size of the data file would still
    be 61000 bytes and the size of the key file would be: <p>
 
<pre>
    ((13 * char) + 1) * 1000
	=  ((13 * 1) + 1) * 1000
	=  14 * 1000
	=  14000 bytes
</pre>
 
    <li><b><i>dbname/table</i>.def</b> - table definition. This file
    contains the table structure definition. <p>
 
    Each field in the table has a 64 byte definition. Using the example
    above, the table has 7 fields so the size of the definition file
    will be: <p>
 
<pre>
    7 * 64  =  448 bytes
</pre>
 
    <li><b><i>dbname/table</i>.stk</b> - table stack. This file keeps
    track of the holes or empty rows in the table. <p>
 
    For every hole in the table, this file will contain a 4 byte
    integer indicating the row number of the hole. It is accessed like
    a stack. When a row is deleted, it's index is appended to the
    file. When an insert is done, the last 4 bytes are
    &quot;popped&quot; off the file and the file is truncated back 4
    bytes. <p>
 
    If the table contains 20 holes, the size of the stack file will be:
    <p>
 
<pre>
    20 * 4  =  80 bytes
</pre>
 
    If the table contains no holes then this file will have zero
    length.
 
</ul>
 
Therefore to calculate the storage requirements for a table, use one of
the following formulae: <p>
 
<dl>
    <dt><b>Tables with a primary key:</b>
 
    <dd>
<pre>
    table_storage_requirements
	=  expected_max_rows *
	    (
		number_of_fields + 1 + total_chars +
		(4 * total_ints) + (8 * total_reals) +
		(size_of_key + 1) +
		(4 * expected_deletion_ratio)
	    ) +
	(total_fields * 64)
</pre>
 
    <dt><b>Tables without a primary key:</b>
 
    <dd>
<pre>
    table_storage_requirements
	=  expected_max_rows *
	    (
		number_of_fields + 1 + total_chars +
		(4 * total_ints) + (8 * total_reals) +
		(4 * expected_deletion_ratio)
	    ) +
	(total_fields * 64)
</pre>
 
</dl>
 
Using the keyed table above, if we expect it to contain a maximum of
10000 rows and we expect to have a 10 percent deletion ratio (that is
at any one time we expect that 10 percent of our rows have been deleted
but not replaced by new rows), we should allow for: <p>
 
<pre>
    10000 *
	(
	    7 + 1 + 13 +
	    (4 * 2) + (8 * 4) +
	    (13 + 1) +
	    (4 * 0.10)
	) +
    (7 * 64)
 
	=  10000 * ( 21 + 8 + 32 + 14 + 0.4) + 448
	=  754448 bytes
</pre>
 
plus a handful of bytes to store file names in directories. <p>
 
<b>Note</b> that this is the maximum storage allocation. Unlike some
other database systems, mSQL only uses disk space when it has data to
add to a table - it does NOT allocate a large block of empty disk space
and then proceed to fill it. If our example only had 1000 rows the
storage requirements would only be 75848 bytes. <p>
 
<hr>
 
<h3><a name="ram">Does <tt>msqld</tt> allocate more RAM to itself as
new databases are added?</a></h3>
 
On Fri, 11 Aug 1995, Andrew Waegel wrote: <p>
 
<pre>
    does msqld allocate more ram to itself as new db's are added? i.e.
    is any part of the database held in ram or does it just access the
    database files directly from disc? I need to do some planning, and
    want to know if I need to plan to get more simms...
</pre>
 
David Hughes replies: <p>
 
<pre>
    If your OS supports mmap() (e.g. Solaris, SunOS, *BSD, BSDI, Linux
    1.3.x, HP-UX &gt;9.x) then the more memory you throw at it the
    better things will get if you are using big databases. The data is
    left on disk but is accessed via the virtual memory subsystem so it
    will be in memory some of the time.
 
    If you are not using mmap() then data is just read from disk as it
    is needed. There's a small buffer in the read code to make things
    faster but that's about it. It doesn't matter how many databases
    you have defined it only uses 1 buffer.
</pre>
 
<hr>
 
<h3><a name="perform">Does performance degrade as the number of
databases increase?</a></h3>
 
On Fri, 11 Aug 1995, Andrew Waegel wrote:
 
<pre>
    Does performance degrade at all as the number of databases
    increases? That is, say a query from database A took n seconds when
    database A was the only one served by msqld. After adding databases
    B, C, D and E, should the database A query take any longer? It
    seems like 'no' from my experience, but...
</pre>
 
David Hughes replies: <p>
 
<pre>
    No. It will degrade if people are hitting the new databases at the
    same time as they are hitting database A though. msqld only handles
    1 query at a time so if 2 queries come in they are queued and
    processed in order.
</pre>
 
<hr>
 
<h3><a name="cursors">Does mSQL support cursors?</a></h3>
 
Pascal Forget replies to a posting by Brian Bartholomew:
 
<pre>
    &gt; To browse the database, I want mSQL to return me the first row
    &gt; in the database, and keep a pointer to it. Then sometime later
    &gt; I can ask it for the second row, and so on.
 
    mSQL does not provide support for cursors.
 
    You'll have to issue a SELECT query each time you want the next
    row. mSQL has no provisions for modifying a result set once it has
    been created. I suggest you add a field containing a unique
    identifier for each row, then fetch the next row using:
 
	SELECT ... FROM mytable where unique_field &gt; last_id LIMIT 1
 
    &gt; How do I express this in sql?
 
    If you find a way to express it, it most certainly won't be in
    standard SQL, as the language has no support for cursors.
 
    &gt; I see that I could add an explicit field that was an arbitrary
    &gt; row number, and query for the current row number +/- 1, but over
    &gt; time with insertions and deletions there would be gaps and the
    &gt; query would break. How is this problem usually solved?
 
    The select statement I gave you won't break even if there are gaps
    in the unique identifiers. You can periodically &quot;compact&quot; the
    numbers if you want.
</pre>
 
<hr>
 
<h3><a name="intl">Does mSQL support different character sets?</a></h3>
 
Patrik Faltstrom writes: <p>
 
<pre>
    Yes, as long as the client that fetches the characters knows what
    character set you are using. I.e. there is no support in mSQL for
    keeping track of the character set name, but mSQL is 8bit clean so
    you can store 8bit characters (in whatever character set).
 
    In Digger, the Whois++ server, we store UNICODE characters by
    encoding them first into UTF-8 which is an 8bit encoding scheme
    described in UNICODE 1.1.
</pre>
 
Neil Bowers <a href="mailto:neilb@cre.canon.co.uk">
&lt;neilb@cre.canon.co.uk&gt;</a> has written a paper on
&quot;<i>Processing Japanese Text with mSQL and Perl</i>&quot;. It is
available from: <p>
 
<pre>
    <a href="http://www.cre.canon.co.uk/~neilb/jmSQL.html">http://www.cre.canon.co.uk/~neilb/jmSQL.html</a>
</pre>
 
<hr>
 
<h3><a name="afs">Why doesn't mSQL work when installed in disk space
served by the Andrew File System (AFS)?</a></h3>
 
Jon Leech <a href="mailto:leech@cs.unc.edu">
&lt;leech@cs.unc.edu&gt;</a> writes: <p>
 
<pre>
    AFS does not support some Unix file system behaviour that's needed
    by mSQL. You cannot create Unix domain sockets in AFS space, and
    mmap() does not work on at least some client-server combinations,
    if not all.
 
    You should install mSQL and its databases in NFS or local Unix
    filesystems.
</pre>
 
Other users have also reported problems when running mSQL from NFS
mounted partitions. To be absolutely safe you should only run
<tt>msqld</tt> from the same machine on which the physical disks
containing the mSQL databases are attached. <p>
 
<hr>
<hr width="75%">
 
<center>
<h2><a name="install-problems">Installation Problems</a></h2>
</center>
 
<h3><a name="irix-username">Under IRIX (SGI) I get problems relating to
my username</a></h3>
 
Colin Surprenant reports: <p>
 
<pre>
    The solution to the problem with using the socket and then nsl
    libraries with NIS in IRIX 5.2 is:
 
    1. Do not link them if they are not needed :) This is the case for
       mSQL.
 
    or
 
    2. link libc BEFORE the socket and the nsl libraries.
 
    For those who didn't know, the problem is that if you use NIS and
    link socket or nsl, the getpwuid() function doesn't work.
</pre>
 
To ensure that the unnecessary libraries are avoided, edit
<tt>targets/<i>your-architecture</i>/site.mm</tt> and change the line
 
<pre>
    EXTRA_LIB=  -lsocket -lnsl -lgen -ldl
</pre>
 
so that it reads: <p>
 
<pre>
    EXTRA_LIB=
</pre>
 
<hr>
 
<h3><a name="osf1-hpux">On OSF/1 or HP-UX I have trouble starting
<tt>msqld</tt> at boot time</a></h3>
 
David-Michael Lincke reports: <p>
 
<pre>
    Looks like the same thing that happens under HP-UX with background
    processes in rc scripts. They are killed off on exit of the ksh
    functions.
 
    Create yourself a wrapper for msqld. In there you do a fork and
    exit the parent process in the child process you do a call to
    setsid() to get rid of the controlling terminal followed by a call
    to execl() to launch msqld. You might also want to close all open
    file descriptors before calling exec.
</pre>
 
An alternative to this approach is to place the following in
<tt>/etc/inittab</tt>
 
<pre>
    msql:3:respawn:/usr/local/Minerva/bin/msqld &lt;/dev/console &gt;/dev/console 2&gt;&amp;1
</pre>
 
This assumes that your mSQL super user is &quot;root&quot;.
<p>
 
The next version of the FAQ will attempt to address this issue in
detail. <p>
 
<hr>
 
<h3><a name="osf1-gcc">Should I use <tt>cc</tt> or <tt>gcc</tt> when
building mSQL on my Dec Alpha running OSF/1?</a></h3>
 
Rick Beebe writes: <p>
 
<pre>
    DEC Alphas running OSF/1 (Digital Unix): The original mSQL docs
    recommended using cc rather than gcc on this platform. In my
    experience this is still good advice. If you have gcc on the
    machine, however, autoconf will find it and default to it. After
    running 'setup' edit site.mm and change 'gcc' to 'cc' and
    'bison -y' to yacc.
</pre>
 
<hr>
 
<h3><a name="linux-mmap">Does mSQL work with Linux
<tt>mmap()</tt>?</a></h3>
 
Version 1.3 or greater of the Linux operating system has full
<tt>mmap()</tt> support. If you're using such a version of Linux, mSQL
will work perfectly well using <tt>mmap()</tt>. <p>
 
If you have an earlier version of Linux you can either upgrade or
ensure that <tt>mmap()</tt> support in mSQL is disabled by running the
'<tt>setup</tt>' program and then editing <p>
 
<pre>
    ./targets/<i>your-architecture</i>/site.mm
</pre>
 
and ensuring the <tt>mmap()</tt> directive reads: <p>
 
<pre>
    MMAP=
</pre>
 
and then recompile the mSQL package. <p>
 
<hr>
 
<h3><a name="hpux-compile">Does mSQL work with HP-UX?</a></h3>
 
Arley Carter (via David Hughes) writes: <p>
 
<pre>
    I just built msql-1.0.10 on hpux 9.05. It appears that you have
    slain the mmap bug. Good job. The only mods I [made] are in site.mm
 
	CC= cc -Ae +O3 +Z
 
    Remove -g flags also or you'll get a lot of annoying messages about
    opt not compatible with debug. The test suites for msql and
    msqlperl ran flawlessly.
</pre>
 
<hr>
 
<h3><a name="hpux-perl">I'm having trouble compiling MsqlPerl with
mSQL under HP-UX</a></h3>
 
Arley Carter writes: <p>
 
<pre>
    This problem has to do with the way HP-UX deals with shared
    libraries.
 
    Ensure the <tt>EXTRA_CFLAGS</tt> option in the file:
 
	./targets/<i>your-architecture</i>/site.mm
 
    reads:
 
	EXTRA_CFLAGS= -Ae +O3 +Z
 
    and recompile mSQL.
 
    The <tt>+Z</tt> option ensures that &quot;position independent
    code&quot; is used when creating object files. For more information
    see the manual pages on your compiler.
</pre>
 
<hr>
 
<h3><a name="sco-unix">How can I install mSQL on a SCO Unix
system?</a></h3>
 
<b>Note:</b> This fix has had some conflicting results. Please let me
<a href="mailto:Peter.Samuel@uniq.com.au">
(&lt;Peter.Samuel@uniq.com.au&gt;)</a> know if it works OR fails. If it
fails, let me know exactly where (if possible) and how you fixed it (if
you did). <p>
 
The following is a summary of the efforts required by Andrew Cash
<a href="mailto:cash_a@sls.co.uk">&lt;cash_a@sls.co.uk&gt;</a>
to install mSQL version 1.0.8 on a SCO Unix system. It should work
perfectly well for version 1.0.9 as well. <p>
 
<ul>
 
    <li> Unpack the source code and create the site dependent files as
    follows:
 
<pre>
    gunzip -c msql-1.0.9.tar.gz | tar xvf -
    cd msql-1.0.9
    make target
    cd targets/<i>your-architecture</i>
    ./setup
	<i>answer questions</i>
</pre>
 
    <li> In the file: <p>
 
<pre>
    ./common/config.h
</pre>
 
    ensure the lines referring to the <tt>sys/select.h</tt> include file are
    commented out as follows: <p>
 
<pre>
    /* Defined if you have sys/select.h  */
    /* #define HAVE_SYS_SELECT_H 1 */
</pre>
 
    <li> <b>Note:</b> This step should only be applied if you use gcc
    and bison. These compilation tools will attempt to use
    <tt>alloca()</tt> instead of <tt>malloc()</tt> which will fail
    under SCO Unix. If you do not use bison, you should skip this
    step.  <p>
 
    SCO Unix doesn't have an <tt>alloca()</tt> library function so
    you'll have to use <tt>malloc()</tt>. You need to generate the
    file: <p>
 
<pre>
    ./msql/msql_yacc.c
</pre>
 
    so run: <p>
 
<pre>
    make
</pre>
 
    until it completes (or fails). Ensure that the <tt>msql_yacc.c</tt>
    file has been created. If it has,
    <a href="#how-to-patch">apply</a>the following patch to the file.
    (Use the &quot;<tt>-l</tt>&quot; option of <tt>patch</tt> to avoid
    any problems with mismatched whitespace. <i>That's an
    &quot;el&quot;, NOT a &quot;one&quot; or an &quot;eye&quot;</i>).
    <p>
 
    This patch ensures that <tt>&lt;malloc.h&gt;</tt> is explicitly
    included and that all references to <tt>alloca()</tt> are changed
    to <tt>malloc()</tt>. <b>Note:</b> This patch has been generated
    based on the output from the <tt>bison</tt> compiler compiler from GNU.
    It should <b>NOT</b> be applied to an <tt>msql_yacc.c</tt>
    that has been generated by <tt>yacc</tt> - such a file already
    uses <tt>malloc()</tt>.
</ul>
 
<pre>
<font size="-1">
*** msql/msql_yacc.c.orig	Fri Jan  5 13:07:02 1996
--- msql/msql_yacc.c	Fri Jan  5 13:09:34 1996
***************
*** 329,362 ****
     Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 
 
- #ifndef alloca
- #ifdef __GNUC__
- #define alloca __builtin_alloca
- #else /* not GNU C.  */
- #if (!defined (__STDC__) &amp;&amp; defined (sparc)) || defined (__sparc__) || defined (__sparc) || defined (__sgi)
- #include &lt;alloca.h&gt;
- #else /* not sparc */
- #if defined (MSDOS) &amp;&amp; !defined (__TURBOC__)
- #include &lt;malloc.h&gt;
- #else /* not MSDOS, or __TURBOC__ */
- #if defined(_AIX)
  #include &lt;malloc.h&gt;
-  #pragma alloca
- #else /* not MSDOS, __TURBOC__, or _AIX */
- #ifdef __hpux
- #ifdef __cplusplus
- extern &quot;C&quot; {
- void *alloca (unsigned int);
- };
- #else /* not __cplusplus */
- void *alloca ();
- #endif /* not __cplusplus */
- #endif /* __hpux */
- #endif /* not _AIX */
- #endif /* not MSDOS, or __TURBOC__ */
- #endif /* not sparc.  */
- #endif /* not GNU C.  */
- #endif /* alloca not defined.  */
 
  /* This is the parser code that is written into each bison parser
    when the %semantic_parser declaration is not specified in the grammar.
--- 329,335 ----
***************
*** 607,618 ****
        yystacksize *= 2;
        if (yystacksize &gt; YYMAXDEPTH)
  	yystacksize = YYMAXDEPTH;
!       yyss = (short *) alloca (yystacksize * sizeof (*yyssp));
        __yy_bcopy ((char *)yyss1, (char *)yyss, size * sizeof (*yyssp));
!       yyvs = (YYSTYPE *) alloca (yystacksize * sizeof (*yyvsp));
        __yy_bcopy ((char *)yyvs1, (char *)yyvs, size * sizeof (*yyvsp));
  #ifdef YYLSP_NEEDED
!       yyls = (YYLTYPE *) alloca (yystacksize * sizeof (*yylsp));
        __yy_bcopy ((char *)yyls1, (char *)yyls, size * sizeof (*yylsp));
  #endif
  #endif /* no yyoverflow */
--- 580,591 ----
        yystacksize *= 2;
        if (yystacksize &gt; YYMAXDEPTH)
  	yystacksize = YYMAXDEPTH;
!       yyss = (short *) malloc (yystacksize * sizeof (*yyssp));
        __yy_bcopy ((char *)yyss1, (char *)yyss, size * sizeof (*yyssp));
!       yyvs = (YYSTYPE *) malloc (yystacksize * sizeof (*yyvsp));
        __yy_bcopy ((char *)yyvs1, (char *)yyvs, size * sizeof (*yyvsp));
  #ifdef YYLSP_NEEDED
!       yyls = (YYLTYPE *) malloc (yystacksize * sizeof (*yylsp));
        __yy_bcopy ((char *)yyls1, (char *)yyls, size * sizeof (*yylsp));
  #endif
  #endif /* no yyoverflow */
</font>
</pre>
 
<ul>
    <li> Rebuild <tt>msqld</tt> with the modified <tt>msql_yacc.c</tt>
    file by running: <p>
 
<pre>
    make
</pre>
 
    <li> If you are NOT going to be running mSQL as root, then you'll
    need to ensure that <tt>msqld</tt> can be installed correctly. Edit
    the makefile: <p>
 
<pre>
    ./msql/Makefile.full
</pre>
 
    and comment out the mode change line for <tt>msqld</tt> as
    follows: <p>
 
<pre>
    # chmod 4750 $(INST_DIR)/bin/msqld
</pre>
 
    Failure to make this change will result in an error during the
    install phase. <p>
 
    <li> Install mSQL by running: <p>
 
<pre>
    make install
</pre>
 
</ul>
 
<hr>
 
<h3><a name="solaris">Why won't my third party applications compile
under Solaris 2.<i>x</i>?</a></h3>
 
Solaris 2.<i>x</i> is an SVR4 version of Unix. When compiling programs
that require network support (this includes any program that uses the
mSQL library), you must explicitly reference the networking libraries
<tt>socket</tt> and <tt>nsl</tt>: <p>
 
<pre>
    gcc -o <i>prog prog.c</i> -I<i>msql_install_dir</i>/include \
	-L<i>msql_install_dir</i>/lib -lmsql <b>-lsocket -lnsl</b>
</pre>
 
<hr>
 
<h3><a name="linux-setup">Why does <tt>setup</tt> fail when building
mSQL on a Linux system?</a></h3>
 
Did you install the Linux kernel sources? <p>
 
If you installed Slackware and didn't get the K series disks then you
will be missing a lot of C header files that you need to compile mSQL.
Go back and grab the kernel sources disks and install them on your box.
<p>
 
<hr>
 
<h3><a name="linux-setup">Why can't I run <tt>setup</tt> under
Linux?</a></h3>
 
This problem involves an interaction between the Linux <tt>setup</tt>
program, the mSQL <tt>setup</tt> program and your <b>$PATH</b>
environment variable. <p>
 
After unpacking the mSQL distribution and running the <tt>make target</tt>
command, the next step in installing mSQL is to change directories to:
<p>
 
<pre>
    targets/<i>target</i>
</pre>
 
and run the <tt>setup</tt> command. If you have <tt>/sbin</tt> in your
<b>$PATH</b> ahead of <tt>.</tt>, or you don't have <tt>.</tt> in your
<b>$PATH</b> at all (which is a good idea if you happen to be
<tt>root</tt>) then your shell will attempt to run the Linux
<tt>setup</tt> program ahead of the mSQL <tt>setup</tt> program. <p>
 
This can be avoided (for all flavours of Unix) by issuing the command
as: <p>
 
<pre>
    ./setup
</pre>
 
This will force your shell to run the <tt>setup</tt> program in the
current working directory, regardless of the contents of your
<b>$PATH</b>. <p>
 
<hr>
 
<h3><a name="bind49">Why do I get errors about
&quot;<tt>sys/bitypes.h</tt>&quot; when compiling under Solaris
2.5?</a></h3>
 
Paul Gregg <a href="mailto:pgregg@tibus.net">
&lt;pgregg@tibus.net&gt;</a> writes: <p>
 
<pre>
    Q: When I try to &quot;make all&quot; mSQL on Solaris (2.5 confirmed) it
    fails with the error:
	../makedepend/makedepend: warning: msqld.c (reading /usr/include/arpa/inet.h,
	line 68): cannot find include file &quot;sys/bitypes.h&quot;
	    not in /usr/include/arpa/sys/bitypes.h
	    not in ../sys/bitypes.h
	    not in /usr/include/sys/bitypes.h
	../makedepend/makedepend: warning: msqld.c (reading /usr/include/arpa/inet.h,
	line 72): cannot find include file "sys/cdefs.h"
	    not in /usr/include/arpa/sys/cdefs.h
	    not in ../sys/cdefs.h
	    not in /usr/include/sys/cdefs.h
 
    A: You have installed BIND 4.9 on your system. Bind replaces your existing
    /usr/include/netdb.h, but forgets to include the compatability includes
    cdefs.h and bitypes.h
 
    Solution: locate the cdefs.h and bitypes.h files in your BIND source code
 
	bind/compat/include/sys/cdefs.h
	bind/compat/include/sys/bitypes.h
 
    and copy these two files to /usr/include/src mSQL should now make
    cleanly.
 
    Credit should be given to Emir Mulabegovic <a href="mailto:mulabego@mcs.anl.gov">&lt;mulabego@mcs.anl.gov&gt;</a> for this
    answer.
 
</pre>
 
<hr>
 
<h3><a name="makedepend">Why do I get errors about
&quot;<tt>POSIX_C_SOURCE</tt>&quot; when compiling under Solaris
2.5?</a></h3>
 
This only occurs when you are compiling mSQL version 1.<i>x</i>. This
version of mSQL ships with a compile time utility called
<tt>makedepend</tt> developed by the X consortium. This utility is
used in the pre compile phase to establish the dependencies of all the
source code used by mSQL. <p>
 
The version of <tt>makedepend</tt> that ships with mSQL version
1.<i>x</i> does not appreciate the ANSI C syntax used by the Solaris
2.5 include files and gives a number of errors such as: <p>
 
 
<pre>
    &quot;strlib.c&quot;:225:  defined(__EXTENSIONS__) || defined(_REENTRANT) ||
	(_POSIX_C_SOURCE - 0 &gt;= 199506L)
				    ^--- expecting )
</pre>
 
These messages can be safely ignored. The actual compilation of the
mSQL source is unaffected. <p>
 
If you wish to avoid seeing these messages you can use the following
procedure to use the <tt>makedepend</tt> that is shipped with
OpenWindows 3.5 instead of the version of <tt>makedepend</tt> shipped
with mSQL version 1.<i>x</i>. <p>
 
<ol>
    <li> <tt>cd msql-1.0.16/src/makegen</tt> <p>
 
    <li> edit both <tt>library.mm</tt> and <tt>object.mm</tt> and
    change all references to
 
<pre>
    @$(TOP)/makedepend/makedepend
</pre>
 
    so that they read: <p>
 
<pre>
    @/usr/openwin/bin/makedepend
</pre>
 
    <li> <tt>cd ../../targets/Solaris-2.5/Sparc</tt> <p>
 
    <li> <tt>make</tt>
 
</ol>
 
This will avoid the error messages listed above but you will now see
the following warning messages: <p>
 
<pre>
    /usr/openwin/bin/makedepend:  ignoring option -a
</pre>
 
These too can be safely ignored. <p>
 
<hr>
 
<h3><a name="sgi-libs">Why do I get errors about &quot;<tt>weak
definitions</tt>&quot; when compiling under IRIX?</a></h3>
 
If you're compiling under Silicon Graphic's IRIX operating system and
see error messages such as: <p>
 
<pre>
    ld: WARNING 134: weak definition of pmap_getport in /usr/lib/libnsl.so
	preempts that weak definition in /usr/lib/libc.so.
</pre>
 
as well as unresolved symbols, you'll need to ensure that unnecessary
libraries are not being used during the compilation. <p>
 
Edit <tt>targets/<i>your-architecture</i>/site.mm</tt>
and change the line
 
<pre>
    EXTRA_LIB=  -lsocket -lnsl -lgen -ldl
</pre>
 
so that it reads: <p>
 
<pre>
    EXTRA_LIB=
</pre>
 
<hr>
 
<h3><a name="aix-ps2">How can I build mSQL version 1.0.16 for an AIX
system running on an IBM PS/2?</a></h3>
 
mSQL assumes that any AIX system will be running on an RS6000
processor. If you are running AIX on an IBM PS/2 the <tt>make
target</tt> will fail. <p>
 
David Schuler
<a href="mailto:schuld@btv.ibm.com">
&lt;schuld@btv.ibm.com&gt;</a> provides the following patch to solve
this problem: <p>
 
<pre>
*** scripts/sys-arch.orig	Fri Mar 28 16:25:49 1997
--- scripts/sys-arch	Fri Mar 28 16:29:19 1997
***************
*** 43,52 ****
 
  if test &quot;$os&quot; = &quot;AIX&quot;
  then
- 	machine=&quot;rs6000&quot;
  	maj=`uname -v`
  	min=`uname -r`
  	rev=&quot;$maj.$min&quot;
  fi
 
  echo &quot;${os}-${rev}-${machine}&quot; | sed 's/ /_/g'
--- 43,57 ----
 
  if test &quot;$os&quot; = &quot;AIX&quot;
  then
  	maj=`uname -v`
  	min=`uname -r`
  	rev=&quot;$maj.$min&quot;
+ 	if test $maj -eq 1
+ 	then
+ 		machine=&quot;i386&quot;
+ 	else
+ 		machine=&quot;rs6000&quot;
+ 	fi
  fi
 
  echo &quot;${os}-${rev}-${machine}&quot; | sed 's/ /_/g'
</pre>
 
To avoid conflicts with mismatched whitespace,
<a href="#how-to-patch">apply</a> this patch using the
&quot;<tt>-l</tt>&quot; option of <tt>patch</tt>.
(<i>That's an &quot;el&quot;, NOT a &quot;one&quot; or an
&quot;eye&quot;</i>). <p>
 
<hr>
 
<h3><a name="apache-auth">How can I make <tt>w3-auth.c</tt> work with
my Apache HTTPD server?</a></h3>
 
You need to patch the Apache server itself. The following patch from
Simon Kershaw <a href="mailto:Simon.Kershaw@Smallworld.co.uk">
&lt;Simon.Kershaw@Smallworld.co.uk&gt;</a> works against Apache 1.2b7:
<p>
 
<pre>
<font size="-1">
*** src/util_script.c~      Mon Feb 10 11:47:24 1997
--- src/util_script.c       Tue Mar 25 15:53:14 1997
***************
*** 174,180 ****
        else if (!strcasecmp (hdrs[i].key, &quot;Content-length&quot;))
            table_set (e, &quot;CONTENT_LENGTH&quot;, hdrs[i].val);
        else if (!strcasecmp (hdrs[i].key, &quot;Authorization&quot;))
!           continue;
        else
            table_set (e, http2env (r-&gt;pool, hdrs[i].key), hdrs[i].val);
      }
--- 174,181 ----
        else if (!strcasecmp (hdrs[i].key, &quot;Content-length&quot;))
            table_set (e, &quot;CONTENT_LENGTH&quot;, hdrs[i].val);
        else if (!strcasecmp (hdrs[i].key, &quot;Authorization&quot;))
!           table_set (e, &quot;HTTP_AUTHORIZATION&quot;, hdrs[i].val);
!           /* continue; */
        else
            table_set (e, http2env (r-&gt;pool, hdrs[i].key), hdrs[i].val);
      }
</font>
</pre>
 
To avoid conflicts with mismatched whitespace,
<a href="#how-to-patch">apply</a> this patch using the
&quot;<tt>-l</tt>&quot; option of <tt>patch</tt>.
(<i>That's an &quot;el&quot;, NOT a &quot;one&quot; or an
&quot;eye&quot;</i>). <p>
 
<hr>
<hr width="75%">
 
<!-- End of Part 2 -->
 
<center>
<h2><a name="runtime-problems">Runtime Problems</a></h2>
</center>
 
<h3><a name="root-user"><tt>msqladmin</tt> will not let me create a
database</a></h3>
 
This one is straight from the manual section on
<a href="#support-progs"><tt>msqladmin</tt></a>. It states that the
only person who is allowed to create a database is the person nominated
as the &quot;root user&quot; during installation. If you indicated that
the database would be running as <tt>root</tt> then you must be
<tt>root</tt> to create a database. <p>
 
If you indicated that it would not be running as <tt>root</tt>, you
would have then nominated a username for the &quot;root user&quot;. In
that case, you must be logged in as the user you nominated before you
can perform admin functions like creation of databases. <p>
 
The manual also states that you can only perform the admin functions of
<a href="#support-progs"><tt>msqladmin</tt></a> (i.e. any function
other than 'version') from the local host. For security reasons you
cannot perform administrative functions in a client/server manner of a
network. <p>
 
<hr>
 
<h3><a name="acl-file">When I start <tt>msqld</tt> it complains about
an ACL file</a></h3>
 
The ACL file is the file that contains Access Control Lists for mSQL.
It is located in the installation directory and is called
<tt>msql.acl</tt> (e.g. <tt>/usr/local/Minerva/msql.acl</tt>). The
warning indicates that <tt>msqld</tt> couldn't locate an ACL file. This
doesn't stop mSQL from operating, it just implies that everyone on
every machine on your network has read/write access to your databases.
<p>
 
A sample ACL file is installed in the installation directory. You could
copy this file to <tt>msql.acl</tt> and edit it to reflect the access
you want to offer to your databases. <p>
 
<hr>
 
<h3><a name="pid-file">When I start <tt>msqld</tt> it complains about a
PID file</a></h3>
 
The PID file is just a file in which
<a href="#support-progs"><tt>msqld</tt></a> writes its process ID. If
it can't write to the PID file, <tt>msqld</tt> will still function
correctly. <p>
 
If you are seeing an error regarding the PID file, then one of the
following could be the cause: <p>
 
<ul>
    <li> The directory in which <tt>msqld</tt> is attempting to create
    the PID file does not exist. <p>
 
    <li> The directory in which <tt>msqld</tt> is attempting to create
    the PID file does not have sufficient permissions.
</ul>
 
During the installation, the <tt>setup</tt> program asked you where to
store the PID file: <p>
 
<pre>
    Top of install tree ? [/usr/local/Minerva]
    Will this installation be running as root ? [y] n
    What username will it run under ?  peters
    Directory for pid file ? [/var/adm]
</pre>
 
You must ensure that this directory exists. The mSQL installation
procedure will NOT create this directory for you. <p>
 
If you did not specify <tt>root</tt> as the mSQL administration user
when you answered the questions: <p>
 
<pre>
    Will this installation be running as root ? [y] n
    What username will it run under ?  peters
</pre>
 
you must ensure that the user you specified has write permissions in
the directory in which <tt>msqld</tt> will store its PID file. <p>
 
Under IRIX 5.3 the <tt>/var/adm</tt> directory can only be written to
by the <tt>root</tt> user, so if your mSQL administration user is NOT
<tt>root</tt> then you'll have to choose some other location such as
<tt>/var/tmp</tt> or <tt>/var/share</tt>. <p>
 
If you need to change the location of this directory, you can either
rerun the <tt>setup</tt> program, or edit the file: <p>
 
<pre>
    ./targets/<i>your_architecture</i>/site.mm
</pre>
 
and change the line: <p>
 
<pre>
    PID_DIR= /var/adm
</pre>
 
to suit your needs. <p>
 
<hr>
 
<h3><a name="recompile">I've just installed the latest version
of mSQL and now my own applications won't work!</a></h3>
 
Whenever you install a new version of mSQL you MUST remember to
recompile any third party applications - including those you developed
yourself - to ensure that they are linked with the new version of
<tt>libmsql.a</tt>. You should also recompile third party applications
after <a href="#how-to-patch">installing any patches</a> to mSQL. <p>
 
Failure to do this will almost certainly guarantee that your
applications will fail at some stage while talking to the new mSQL
database server. You may also miss out on some new feature provided by
the new mSQL API. <p>
 
<hr>
 
<h3><a name="setuid">Access control doesn't work with my setuid
applications</a></h3>
 
<b>Note:</b> This only applies to mSQL version 1.0.10 and previous
versions. mSQL version 1.0.12 (and above) does NOT suffer from this
problem. This section will be removed in future versions of the FAQ. It
is included in this release to accommodate the transitional period
following the recent release of version mSQL 1.0.12. <p>
 
Consider the following scenario: <p>
 
<ul>
    <li> User <tt>peters</tt> is the ONLY user given read and write
    access to the database <i>test</i> by the appropriate additions to
    the <tt>msql.acl</tt> file. <p>
 
    <li> An application is created to manipulate the <i>test</i>
    database and is saved such that it is a <i>setuid peters</i>
    application: <p>
 
<pre>
    -rwsr-xr-x   1 peters   db         24576 Nov 13  1995 db_app
</pre>
 
    <li> User <tt>markp</tt> wishes to use the <i>test</i> database
    using the <tt>db_app</tt> application. Instead of the <i>setuid</i>
    nature of the application allowing <tt>markp</tt> access to the
    database, all that happens is the error message: <p>
 
<pre>
    Access to database denied
</pre>
 
</ul>
 
This occurs because of the way in which mSQL passes user information to
the database server. It uses the <i>getuid()</i> system call which
returns the real user id of the calling process rather than the
effective user id. <p>
 
mSQL version 2 will have radically different security mechanisms. <p>
 
<hr>
 
<h3><a name="address-in-use">Why do I see an &quot;<tt>Address already
in use</tt>&quot; error message when I attempt to start
<tt>msqld</tt>?</a></h3>
 
If you're running Linux or SCO Unix, this message can be seen if you
killed the <a href="#support-progs"><tt>msqld</tt></a> process WITHOUT
executing: <p>
 
<pre>
    msqladmin shutdown
</pre>
 
The TCP/IP port will remain bound for about 90 seconds or so. After
this time the port should be available and <tt>msqld</tt> can be
started successfully. <p>
 
Another possibility to consider is that something is already using the
TCP/IP port that <a href="#support-progs"><tt>msqld</tt></a> is trying
to use. For a default installation these port numbers are 1112 for a
root user or 4333 for an ordinary user. In this case <i>user</i> means
the name of the user you entered when answering the setup question(s):
<p>
 
<pre>
    Will this installation be running as root ?
    What username will it run under ?
</pre>
 
There are a number of ways you can check for something using the TCP/IP
port: <p>
 
<ul>
    <li><b>msql</b> <p>
 
    If the mSQL monitor program <tt>msql</tt> can connect to the mSQL
    database server <tt>msqld</tt> then you KNOW that the database
    server is already running. <p>
 
    <li><b>Telnet</b> <p>
 
    Telnet to the database server and specify the mSQL TCP/IP port
    number using one of the following commands: <p>
 
<pre>
    telnet <i>dbhost</i> 1112
</pre>
 
or
 
<pre>
    telnet <i>dbhost</i> 4333
</pre>
 
    You'll see the following types of messages: <p>
 
    <ul>
 
    <li> Nothing is using that port or something is using the port
	exclusively: <p>
 
<pre>
    Trying 127.0.0.1 ...
    telnet: connect: Connection refused
    telnet&gt;
</pre>
 
    <li><tt>msqld</tt> version 1.0.10 running protocol version 6 is
    using the port: <p>
 
<pre>
    Trying 127.0.0.1 ...
    Connected to localhost.
    Escape character is '^]'.
 
    0:6:1.0.10
</pre>
 
    <li> Something else is using the port (or a client process is still
    communicating with a running <tt>msqld</tt> process: <p>
 
<pre>
    Trying 127.0.0.1 ...
    Connected to localhost.
    Escape character is '^]'.
</pre>
 
    </ul>
 
    <li><b>netstat</b> <p>
 
    If your operating system has the <tt>netstat</tt> command, you can
    use it to display the contents of various network related data
    structures in various formats, depending on the options you
    select.
 
    Some of the options that may be helpful are: <p>
 
    <ul>
 
    <li> To determine if something is listening on port 4333, use: <p>
 
<pre>
    netstat -a | grep 4333
</pre>
 
    If you see output similar to:
 
<pre>
      *.4333	*.*	0	0	0	0	LISTEN
</pre>
 
    then something is using that port. <p>
 
    <li> To see if <tt>msqld</tt> is using the port, use: <p>
 
<pre>
    netstat -f unix
</pre>
 
    Output similar to the following will indicate that <tt>msqld</tt>
    is already running: <p>
 
<pre>
    Active UNIX domain sockets
    Address  Type          Vnode     Conn Addr
    fcf8bca8 stream-ord      231        0 /tmp/msql.sock
</pre>
 
    This may not work for all operating systems. - the above examples
    were taken from a Solaris 2.4 system. Variations on this command
    include: <p>
 
<pre>
    netstat -f inet
</pre>
 
    or
 
<pre>
    netstat -f local
</pre>
 
    </ul>
 
</ul>
 
If you found something using the TCP/IP port that <tt>msqld</tt> would
like to use, examine the output of your operating system's <tt>ps</tt>
command to check if you already have a running <tt>msqld</tt> process.
If you have then shut it down it by executing: <p>
 
<pre>
    msqladmin shutdown
</pre>
 
If you don't have a running <tt>msqld</tt> process then something else
may be using the port that <tt>msqld</tt> is trying to use. Examine
<tt>/etc/inetd.conf</tt> and <tt>/etc/services</tt> (or the services
NIS map if you're running NIS) to see if anything else is using the
port. The output from one of the <tt>netstat</tt> commands listed
above may be helpful. <p>
 
If you find such a program you have two options: <p>
 
<ol>
    <li> Change the port number the other program is using so that it
    doesn't conflict with mSQL <p>
 
    <li> Ensure that mSQL uses a different port number. This can be done
    by either: <p>
 
	<ul>
	    <li> starting <tt>msqld</tt> with the environment variable
	    MSQL_TCP_PORT set to another port number. You'll also have
	    to ensure that client applications use the new port number
	    too. <p>
 
	    <li> modifying the <tt>./src/common/site.h</tt> file and
	    changing the section
 
<pre>
    /*
    ** TCP port for the MSQL daemon
    */
 
    #ifdef ROOT_EXEC
    #define MSQL_PORT       1112
    #else
    #define MSQL_PORT       4333
    #endif
</pre>
 
	    to suit your needs. Then recompile and reinstall mSQL.
	</ul>
</ol>
 
<hr>
 
<h3><a name="bind-permissions">Why do I see a &quot;<tt>Can't start
server : UNIX Bind : Permission denied</tt>&quot; error message when I
attempt to start <tt>msqld</tt>?</a></h3>
 
This usually occurs because of a permission problem involving the Unix
domain socket used by
<tt>msqld</tt> when it communicates with processes on the same host as
itself. <p>
 
Under mSQL version 1.<i>x</i> this file is typically
<tt>/dev/msql</tt> or <tt>/tmp/msql.sock</tt> while under mSQL
version 2.<i>x</i> it is typically <tt>%I/msql2.sock</tt> where
<tt>%I</tt> is a reference to the directory specified by
<tt>Inst_Dir</tt> in the mSQL configuration file. <p>
 
These permission problems usually involve the socket itself or the
directory in which it resides. The general rule here is that the
directory should have read and write permission for the mSQL root user,
and the socket (or named pipe on some operating systems) should be
owned by the mSQL root user. (<b>Note:</b> The mSQL root user is the
user you specified during the setup stage during the installation of
mSQL version 1.<i>x</i> or the user specified in the
<tt>msql.conf</tt> file for mSQL version 2.<i>x</i>). <p>
 
Some operating systems, such as HP-UX 10.<i>x</i>, ship with
<tt>dr-xr-xr-x</tt> permissions on <tt>/dev</tt> which makes it
difficult for <tt>msqld</tt> to write to the socket. In these
circumstances a better place for the socket would be
<tt>/var/msql</tt>. <p>
 
The installation instructions for mSQL 2.<i>x</i> state: <p>
 
<pre>
    8.      Ensure that the installation directory (and everything under it)
	    is owned by the user that mSQL will run as (e.g. the msql user).
	    To do this, issue a command like "chown -R msql /usr/local/Hughes".
	    Substitute the correct username and installation directory if
	    you are running a non-default installation.
</pre>
 
If you decide to place the mSQL Unix domain socket in one of the
temporary directories such as <tt>/tmp</tt> <tt>/usr/tmp</tt> or
<tt>/var/tmp</tt> please read the section in this FAQ titled
<a href="#socket-location">&quot;Where did my mSQL socket file
go?&quot;</a> <p>
 
You should also read the section in this FAQ titled
<a href="#socket-location">&quot;Where did my mSQL socket file
go?&quot;</a> if your operating system does not ship with write
permission on <tt>/dev</tt>. <p>
 
<hr>
 
 
<h3><a name="tmpdir">Why do I see &quot;<tt>Couldn't create temporary
table</tt>&quot; error messages?</a></h3>
 
mSQL uses a temporary area to store the results of table joins while
it is processing a query. This area lives in the directory: <p>
 
<pre>
    msqldb/.tmp
</pre>
 
under the directory in which you elected to install mSQL. <p>
 
If the <tt>.tmp</tt> directory does not exist, or doesn't have
permissions that allow the mSQL super user to create files then you'll
see the "<tt>Couldn't create temporary table</tt>" message. <p>
 
To fix this, check for the existence of the directory and ensure
that it has permissions:
 
<pre>
    drwxr-xr-x   2 <i>owner</i>  <i>group</i>      512 Jan 28 16:10 .tmp
</pre>
 
where <i>owner</i> is the user name of the mSQL super user and
<i>group</i> is whatever is appropriate to the way in which you want to
run your system (in other words it isn't terribly important in this
case). <p>
 
If you're going to do really big table joins then have a look at the
section in the FAQ titled <a href="#tmp-space">&quot;How can I avoid
running out of space when doing certain complex table joins?&quot;</a>.
<p>
 
<hr>
 
<h3><a name="tmp-space">How can I avoid running out of space when doing
certain complex table joins?</a></h3>
 
Some mSQL operations involving table joins can consume large amounts of
temporary disk space. You can change the location of this temporary
storage area using the following procedure: <p>
 
<ul>
    <li> Shutdown the mSQL database server. <p>
 
<pre>
    msqladmin shutdown
</pre>
 
    <li> Remove the <tt>msqldb/.tmp</tt> directory from your mSQL
    installation directory. For example, if you installed mSQL in
    <tt>/usr/local/Minerva</tt> run the following command: <p>
 
<pre>
    rm -rf /usr/local/Minerva/msqldb/.tmp
</pre>
 
    <li> Create a symbolic link from a directory with lots of space to
    the <tt>msqldb/.tmp</tt> directory: <p>
 
<pre>
    ln -s /<i>lots</i>/<i>of</i>/<i>space</i> /usr/local/Minerva/msqldb/.tmp
</pre>
 
    <li> Restart the mSQL database server.
 
<pre>
    msqld &
</pre>
 
</ul>
 
<hr>
 
<h3><a name="bzero-missing"><tt>msqld</tt> is suddenly dumping core and
complaining about <tt>bzero()</tt></a></h3>
 
Rasmus Lerdorf writes: <p>
 
<pre>
    This looks to me like you have compiled your msqld binary on a
    machine with bzero() in your libc but you are running it on a
    machine that does not have the bzero() function in its libc. Could
    it be that you compiled on a Solaris 2.5 box, but you are running
    it on a Solaris 2.4 machine? Sun made the brilliant decision to add
    bzero, bcopy and rindex to the Solaris 2.5 libc which means that
    Solaris 2.5 and 2.4 are not completely binary compatible anymore.
    When you compile msql on Solaris 2.5 it will see that bzero and
    bcopy are available and thus try to use them.
 
    The fix is to either link your msqld statically with the Solaris
    2.5 libc, or perhaps more preferable, recompile msql to not use the
    silly bzero, bcopy, bcmp and rindex functions.
 
    Edit common/config.h and make sure you do not have HAVE_BCOPY and
    HAVE_RINDEX defined. If you do, comment out these definitions and
    recompile. The resulting binary should then run on both target
    machines.
</pre>
 
The exact error message you'll see when running a Solaris 2.5 compiled
<tt>msqld</tt> on a Solaris 2.4 machine is: <p>
 
<pre>
    peters@wheel[710] ./msqld
 
    mSQL Server 1.0.10 starting ...
 
    ld.so.1: ./msqld: fatal: relocation error: symbol not found: bzero:
    referenced in ./msqld
    Killed
</pre>
 
<hr>
 
<h3><a name="bsdlib">Why does <tt>relshow</tt> drop the first two
characters from its output?</a></h3>
 
Desmond writes: <p>
 
<pre>
    Wonder if anyone encountered this weird display while using relshow?
 
    relshow bookmarks
 
    Database = bookmarks
 
    +---------------------+
    |       Table         |
    +---------------------+
    | okmarks             |
    | st                  |
    +---------------------+
 
    notice that the first two letters of the table names are missing.
    Please help. Thanks in advance!
</pre>
 
David Hughes replies: <p>
 
<pre>
    I've seen this on Solaris if you link against the oh so broken BSD
    compatibility library (libbsd.a). Please make sure that libbsd.a
    isn't mentioned in your site.mm file.
</pre>
 
And from the comp.unix.solaris FAQ: <p>
 
<pre>
    6.19) Why doesn't readdir work?  It chops the first two characters of
	all filenames.
 
	You're probably linking with libucb and didn't read question
	6.18. Readdir in libucb.so wants you to include sys/dir.h,
	but many SunOS 4.1.x programs included &lt;dirent.h&gt;,
	consequently, you're mixing native &lt;dirent.h&gt; struct dirent
	with libucb readdir(). The symptom of this mixup is that the
	first two characters of each filename are missing. Make sure
	you use the native compiler (default /opt/SUNWspro/bin/cc,
	which may not be in your PATH), and not /usr/ucb/cc.
</pre>
 
If you haven't shelled out the cash for Sun's compiler you can
substitute <tt>gcc</tt> for <tt>/opt/SUNWspro/bin/cc</tt> above. <p>
 
<hr>
 
<h3><a name="msqladmin">Why won't <tt>msqladmin</tt> work?</a></h3>
 
When running <tt>msqladmin</tt> you may occasionally see error
messages: <p>
 
<pre>
    mSQL command failed!
    Server error = Permission denied
</pre>
 
<tt>msqladmin</tt> has the following restrictions: <p>
 
<ol>
    <li> it can only be run by the msql root user. That is the user you
    specified when you compiled msql. This is a security feature.
    <p>
 
    <li> it can only be run on the machine on which msqld is running -
    it will NOT work via tcp/ip connections. This is a security
    feature.
 
</ol>
 
The exception to the above restrictions is <p>
 
<pre>
    msqladmin version
</pre>
 
this command can be run by any user, even over a network connection.
<p>
 
<hr>
 
<h3><a name="aix-compress">Why won't mSQL work with the compressed file
system under AIX?</a></h3>
 
It seems that mSQL will corrupt its database files if they are stored
using the compressed file system running under AIX. The reasons are not
yet clear but may be related to AIX's implementation of <tt>mmap()</tt>.
<p>
 
The corruption can be avoided by ensuring that the database files are
not stored in a compressed file system. <p>
 
<hr>
 
<h3><a name="hpux-index">Why do I see &quot;<tt>Protocol mismatch</tt>"
errors under HP-UX?</a></h3>
 
The following error message: <p>
 
<pre>
    Protocol mismatch. Server Version = 0 Client Version = 6
</pre>
 
may indicate that the version of <tt>index()</tt> being used is
broken. This is true with some implementations of HP-UX 9.<i>x</i>. <p>
 
mSQL's <tt>autoconf</tt> build procedure will use <tt>index()</tt> if
it finds it. The fix is to modify: <p>
 
<pre>
    ./targets/<i>your_architecture</i>/common/config.h
</pre>
 
and remove (or comment out) the line that reads: <p>
 
<pre>
    #define HAVE_RINDEX 1
</pre>
 
Then recompile mSQL. <p>
 
<hr>
 
<h3><a name="machten">Why do I see <tt>&quot;Can't start server :  UNIX
Bind : Invalid argument&quot;</tt> errors running <tt>msqld</tt> under
a MachTen BSD Unix system?</a></h3>
 
Mark Murphy <a href="mailto:markm@desktop.tyrell.com">
&lt;markm@desktop.tyrell.com&gt;</a> writes: <p>
 
<pre>
    At 12:10 PM 9/17/96, I wrote:
    &gt;I'm trying to install mSQL on a MachTen BSD unix system and am having
    &gt;problems. The compile and installation went great (with only a warning
    &gt;that it could not detect 'uname' on the system). But when I tried to start
    &gt;the server from root with:
    &gt;
    &gt;/usr/local/Minerva/bin/msqld&amp;
    &gt;
    &gt;I get the following:
    &gt;
    &gt;Can't start server : UNIX Bind : Invalid argument
    &gt;
    &gt;Anyone have any suggestions on what I am doing wrong or what is missing?
 
 
    Thought I'd let everyone know what the problem was in case anyone else is
    thinking of using MachTen BSD Unix
 
    There's a bug in the &lt;sys/un.h&gt; file.  It reads:
 
    /*
     * Definitions for UNIX IPC domain.
     */
    struct  sockaddr_un {
	    short   sun_family;             /* AF_UNIX */
	    char    sun_path[108];          /* path name (gag) */
    };
 
 
    And it should be:
 
    /*
     * Definitions for UNIX IPC domain.
     */
    struct  sockaddr_un {
	    short   sun_family;             /* AF_UNIX */
	    char    sun_path[106];          /* path name (gag) */
    };
 
 
    It seems sun_path was a little too long and the call to 'bind' in msqld
    does a check on the parameter's structure size... thus giving an invalid
    argument error.
 
    On another note....
 
    After I got that bug fixed, the server started, but when exercising the
    tests, msql crashed MachTen!  YES... I mean crash!  So much for protected
    memory!
 
    But the real problem is yet another bug with MachTen.
 
    It seems MachTen's mmap function &quot;has not been fully
    tested&quot; (as one tech support person finally admitted). So
    while mmap exists, it really doesn't work.
 
    Borrowing from the msql FAQ:
 
	&gt;Version 1.3 or greater of the Linux operating system has full mmap support.
	&gt;If you're using such a version of Linux, mSQL will work perfectly well using
	&gt;mmap.
	&gt;
	&gt;If you have an earlier version of Linux you can either upgrade or ensure that
	&gt;mmap support in mSQL is disabled by running the 'setup' program and then
	&gt;editing
	&gt;
	&gt;    ./<i>targets/your-architecture</i>/site.mm
	&gt;
	&gt;and ensuring the mmap directive reads:
	&gt;
	&gt;    MMAP=
	&gt;
	&gt;and then recompile the mSQL package.
 
    This also works for MachTen since it seems mmap is &quot;not fully tested&quot;...
    and in reality it damn well crashed the machine.
 
    These two issues took two full days to resolve.... with calls to Tenon tech
    support in the morning, suggestions that were nowhere near fixing the
    problem during the day... and the &quot;real&quot; answers coming near 5pm when tech
    support closes... that these are bugs in MachTen.
 
    While their tech support staff was always nice and tried to be very
    helpful, I wish they would have been able to answer my questions when I
    first called.  It would have saved me two days of work on these issues.
 
    They had me try all kinds of things until the call was elevated to another
    tech (he doesn't get in until late afternoon) who had the answers... it's a
    MachTen bug!
 
    Very frustrating.
 
    However on a positive point, MachTen has been running very good as my
    internet server for the past six months or so without a glitch.
 
</pre>
 
<hr>
 
<h3><a name="socket-location">Where did my mSQL socket file
go?</a></h3>
 
Consider these two scenarios: <p>
 
<ol>
 
    <li> Your machine has just been rebooted and suddenly applications
    running on the same host cannot connect to <tt>msqld</tt>. After
    some investigation you discover that <tt>/tmp/msql.sock</tt> has
    been removed - why? <p>
 
    <li> Your mSQL database server has been running fine for a week or
    so and suddenly applications running on the same host cannot
    connect to <tt>msqld</tt>. After some investigation you discover
    that <tt>/tmp/msql.sock</tt> has been removed - why?
 
</ol>
<p>
 
You've probably been bitten by an operating feature and/or some good
system administration practise. Firstly some background. <p>
 
The two classic temporary areas under most flavours of Unix are
<tt>/tmp</tt> and <tt>/usr/tmp</tt> (<tt>/var/tmp</tt> in some
systems). They are usually on different file systems but don't have to
be. <p>
 
Under some versions of Unix, <tt>/tmp</tt> is cleared whenever the
system is rebooted. Other versions don't do this. Most versions of Unix
do NOT clear <tt>/usr/tmp</tt> on reboot. <p>
 
Some operating systems - such as Solaris 2.<i>x</i> and SunOS
4.<i>x</i> under certain configurations - take advantage of unused
portions of the swap partition to provide <tt>/tmp</tt>. (This happens
by default under Solaris 2.<i>x</i> and can be turned on using the
TMPFS file system under SunOS 4.<i>x</i>).  Because this area is not a
real file system <tt>/tmp</tt> only has a transient lifetime and is
guaranteed to be empty after each reboot. <p>
 
While the system is running, some versions of Unix regularly clear
<tt>/tmp</tt> and possibly <tt>/usr/tmp</tt> of files older than some
arbitrary period (usually a few days or more). This procedure is
handled by <tt>root</tt> via <tt>cron</tt>. Some versions of Unix are
shipped with a <tt>root</tt> crontab to perform this cleanup, others
aren't.  Often the entry in crontab is draconian enough to remove
everything older than the required period - including directories,
named pipes, sockets etc.  Other entries will only remove files. In is
not unusual for system administration staff to add a similar entry to
<tt>root</tt>'s crontab if it doesn't already exist. <p>
 
So what this boils down to is that files in <tt>/tmp</tt> and
<tt>/usr/tmp</tt> (<tt>/var/tmp</tt>) are <b>TEMPORARY</b> and should
be treated as such. If you need to create a socket for mSQL you should
follow these simple rules: <p>
 
<ol>
    <li> If <tt>root</tt> is your mSQL master user, the socket should
    go in <tt>/dev</tt>. This happens by default under both mSQL version
    1.<i>x</i> and 2.<i>x</i>. <p>
 
    However some operating systems such as HP-UX 10.<i>x</i> ship with
    <tt>dr-xr-xr-x</tt> permissions on <tt>/dev</tt>. In this case,
    you should choose a different location entirely for the mSQL Unix
    domain socket. <tt>/var/msql</tt> is one possible choice. <p>
 
    For both mSQL version 1.<i>x</i> and 2.<i>x</i> this can be
    accomplished by editing: <p>
 
<pre>
    ./targets/<i>your-architecture</i>/site.h
</pre>
 
    and modifying the value of
 
<pre>
    #define MSQL_UNIX_ADDR  &quot;/dev/msql&quot;
</pre>
 
    for mSQL version 1.<i>x</i>, or
 
<pre>
    #define MSQL_UNIX_ADDR  &quot;/dev/msql2&quot;
</pre>
 
    for mSQL version 2.<i>x</i>. <p>
 
    Then re-compile. <p>
 
    <li> If <tt>root</tt> is not your mSQL master user, the socket
    should go in a directory that the mSQL master user can read/write
    but not one of the temporary directories <tt>/tmp</tt>,
    <tt>/usr/tmp</tt> or <tt>/var/tmp</tt>. <p>
 
    <ul>
 
	<li> For both mSQL version 1.<i>x</i> and 2.<i>x</i> this can
	be accomplished by editing
 
<pre>
    ./targets/<i>your-architecture</i>/site.h
</pre>
 
	and modifying the value of
 
<pre>
    #define MSQL_UNIX_ADDR  &quot;/tmp/msql.sock&quot;
</pre>
 
	Then re-compile. <p>
 
	<li> For mSQL version 1.<i>x</i> you can also start
	<tt>msqld</tt> with the environment variable
	<b>MSQL_UNIX_PORT</b> set to the new location of the mSQL
	socket. There is no need to recompile in this case. You will
	have to ensure that this environment variable is set for ALL
	programs that wish to communicate with <tt>msqld</tt> -
	including <tt>msql</tt>, <tt>msqladmin</tt>, <tt>relshow</tt>,
	<tt>MsqlPerl</tt>, etc etc. <p>
 
	<li> For mSQL 2.<i>x</i> you can also edit the mSQL config file
 
<pre>
    msql.conf
</pre>
 
	and change the value of
 
<pre>
    UNIX_PORT
</pre>
 
	By default this is <tt>%I/msql2.sock</tt> where <tt>%I</tt> is
	the directory in which you installed mSQL 2.<i>x</i>. There is
	no need to recompile in this case.
 
    </ul>
 
</ol>
<p>
 
<hr>
 
<h3><a name="slow-connections">Why does it take such a long time to
connect to the mSQL database server?</a></h3>
 
Sometimes you may have difficulty in connecting to the mSQL database
server - <tt>msqld</tt>. <p>
 
<tt>msqld</tt> is a single threaded application and therefore can only
deal with one SQL query at a time. If <tt>msqld</tt> is busy servicing
an SQL query you may not be able to make another connection until the
SQL query has finished. mSQL version 1.<i>x</i> is extremely inefficient
in performing table joins. So an SQL query with only a single table
join may take many minutes to complete. <p>
 
There is no solution to this problem in mSQL version 1.<i>x</i>. The
only possible workaround is to ensure that your database schema is
designed in such a way as to avoid table joins at all costs. <p>
 
Table joins are much more efficient under mSQL version 2.<i>x</i> so
the problem will rarely be encountered. <p>
 
<hr>
 
<h3><a name="beta-expire">Why do I see <tt>&quot;This Beta test release
expires in 57 days&quot;</tt> when I start <tt>msql2d</tt>?</a></h3>
 
The current releases of mSQL version 2.<i>x</i> are beta versions. By
their very nature, beta versions are still in development. Hughes
Technologies has released the beta versions in an attempt to leverage
off the existing user base in order to do final testing and bug
fixing. <p>
 
The expiration messages are reminders to the users of the beta version
that Hughes Technologies expects to have either a new beta version or
a completed version released within the time limit suggested by the
message. <p>
 
If the time limit has elapsed, <tt>msql2d</tt> will cease to operate
and you will be presented with a message: <p>
 
<pre>
    This Beta test release has expired!
    Check out www.Hughes.com.au for a new release!
</pre>
 
This site has a few problems at present. Use
<a href="http://hughes.com.au">http://hughes.com.au</a>
instead. <p>
 
If a new version of mSQL 2.<i>x</i> (beta or otherwise) is not
available after the time limit has expired, it is a simple matter
to disable this feature using one of the patches listed below. Once
you have applied the patch you MUST recompile and re-install. There is
no need to recompile any third party applications as this patch only
affects the operations of the database server. <p>
 
To avoid conflicts with mismatched whitespace,
<a href="#how-to-patch">apply</a> the patch using the
&quot;<tt>-l</tt>&quot; option of <tt>patch</tt>.
(<i>That's an &quot;el&quot;, NOT a &quot;one&quot; or an
&quot;eye&quot;</i>). <p>
 
<ul>
    <li> mSQL 2.0B1 <p>
 
<pre>
<font size="-1">
*** src/msql/msqld.c.orig	Mon Dec 16 19:45:33 1996
--- src/msql/msqld.c	Wed Apr 30 10:08:49 1997
***************
*** 583,589 ****
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (850703170 + 91*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Snapshot release has expired!\n&quot;);
--- 583,589 ----
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (999999999 + 91*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Snapshot release has expired!\n&quot;);
</font>
</pre>
 
    <li> mSQL 2.0B2 <p>
 
<pre>
<font size="-1">
*** src/msql/msqld.c.orig	Wed Apr 30 09:57:54 1997
--- src/msql/msqld.c	Wed Apr 30 10:11:37 1997
***************
*** 591,597 ****
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (850703170 + 91*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Snapshot release has expired!\n&quot;);
--- 591,597 ----
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (999999999 + 91*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Snapshot release has expired!\n&quot;);
</font>
</pre>
 
    <li> mSQL 2.0B3 <p>
 
<pre>
<font size="-1">
*** src/msql/msqld.c.orig	Wed Apr 30 09:57:54 1997
--- src/msql/msqld.c	Wed Apr 30 10:11:48 1997
***************
*** 595,601 ****
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (851999170 + 91*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Snapshot release has expired!\n&quot;);
--- 595,601 ----
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (999999999 + 91*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Snapshot release has expired!\n&quot;);
</font>
</pre>
 
    <li> mSQL 2.0B4 <p>
 
<pre>
<font size="-1">
*** src/msql/msqld.c.orig	Wed Apr 30 09:57:54 1997
--- src/msql/msqld.c	Wed Apr 30 10:11:58 1997
***************
*** 600,606 ****
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (854763970 + 91*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Snapshot release has expired!\n&quot;);
--- 600,606 ----
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (999999999 + 91*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Snapshot release has expired!\n&quot;);
</font>
</pre>
 
    <li> mSQL 2.0B5 <p>
 
<pre>
<font size="-1">
*** src/msql/msqld.c.orig	Wed Apr 30 09:57:54 1997
--- src/msql/msqld.c	Wed Apr 30 10:12:08 1997
***************
*** 651,657 ****
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (858088800 + 60*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Beta test release has expired!\n&quot;);
--- 651,657 ----
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (999999999 + 60*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Beta test release has expired!\n&quot;);
</font>
</pre>
 
    <li> mSQL 2.0B6 <p>
 
<pre>
<font size="-1">
*** src/msql/msqld.c.orig	Wed Apr 30 09:57:54 1997
--- src/msql/msqld.c	Wed Apr 30 10:12:17 1997
***************
*** 696,702 ****
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (858088800 + 60*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Beta test release has expired!\n&quot;);
--- 696,702 ----
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (999999999 + 60*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Beta test release has expired!\n&quot;);
</font>
</pre>
 
    <li> mSQL 2.0B7 <p>
 
<pre>
<font size="-1">
*** src/msql/msqld.c	Mon May 12 17:35:02 1997
--- src/msql/msqld.c.orig	Thu May  8 11:54:23 1997
***************
*** 696,702 ****
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (999999999 + 60*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Beta test release has expired!\n&quot;);
--- 696,702 ----
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (863060000 + 60*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Beta test release has expired!\n&quot;);
</font>
</pre>
 
    <li> mSQL 2.0B7.1 <p>
 
<pre>
<font size="-1">
*** src/msql/msqld.c	Mon May 12 17:35:08 1997
--- src/msql/msqld.c.orig	Thu May  8 11:54:23 1997
***************
*** 696,702 ****
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (999999999 + 60*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Beta test release has expired!\n&quot;);
--- 696,702 ----
  	chdir(msqlHomeDir);
  	cleanTmpDir();
 
! 	timeRemain = (863060000 + 60*24*60*60) - time(NULL);
  	if (timeRemain &lt; 0)
  	{
  		printf(&quot;\nThis Beta test release has expired!\n&quot;);
</font>
</pre>
 
</ul>
 
<hr>
 
<h3><a name="acl-ips">Can I use IP addresses in the mSQL ACL file?</a></h3>
 
Yes. <p>
 
I. Chazakis <a href="mailto:ioannis@acropolis.net">
&lt;ioannis@acropolis.net&gt;</a> reports: <p>
 
<pre>
    I've been using ip addresses in acl files since day one, without
    any problems in version 1.xx, so I guess it should work for you as
    well.
</pre>
 
<hr>
 
<h3><a name="protocol-mismatch">What does the &quot;Protocol mismatch.
Server Version = 22 Client Version = 6&quot; error mean?</a></h3>
 
Messages similar to this mean that your client software is attempting
to connect to an mSQL database server which is running a different
version of the mSQL protocol. This could be a result of an old client
attempting to talk to a new server or a new client attempting to talk
to an old server. <p>
 
The numbers mentioned in the message give an indication of which
versions of mSQL each end of the session is using. In the example
given, the client was compiled with mSQL version 1.<i>x</i>
(specifically 1.0.16) because the protocol version is less than 10, and
the server is mSQL version 2.<i>x</i> because the protocol version is
greater than 20. <p>
 
This problem can be rectified by recompiling your third party mSQL code
(that is any client code that did NOT come with the mSQL distribution,
including software such as MsqlPerl etc) against the latest version of
mSQL available to you. <p>
 
Also see the section in the FAQ titled <a href="#recompile">&quot;I've
just installed the latest version of mSQL and now my own applications
won't work!&quot;</a>. <p>
 
<hr>
 
<h3><a name="query-ordering">Why does mSQL take longer to return the
same results when I change the order of my query?</a></h3>
 
Michael Wiedmer <a href="mailto:mwiedmer@eso.org">
&lt;mwiedmer@eso.org&gt;</a> writes: <p>
 
<pre>
    mSQL behaves like a charm - mostly. That is, depending on the ordering
    of the tables in the FROM clause of a SELECT query, mSQL will return
    the results in different time scales - if at all!
 
    These two SELECT queries only differ in the FROM clause. You will
    notice that the &quot;component&quot; table comes last and first, respectively.
 
      SELECT component.location
	FROM instrument, has_dpr, data_product, component
       WHERE instrument.ins_id LIKE 'SUSI'
	 AND instrument.ins_id = has_dpr.comp_id
	 AND has_dpr.dpr_id = data_product.dp_id
	 AND component.comp_id = data_product.comp_id
	 AND data_product.type = 'PRG'
 
 
      SELECT component.location
	FROM component, instrument, has_dpr, data_product
       WHERE instrument.ins_id LIKE 'SUSI'
	 AND instrument.ins_id = has_dpr.comp_id
	 AND has_dpr.dpr_id = data_product.dp_id
	 AND component.comp_id = data_product.comp_id
	 AND data_product.type = 'PRG'
 
    In version 1.0.16, the latter fails miserably (it returns after 5
    minutes or so), whereas the former completes the query correctly
    within seconds.
 
    In version 2.0 beta 6, the latter returns after 30 seconds, whereas
    the former completes the query wrongly!!
 
    Could you tell me whether this is a known feature and whether there
    are any workarounds or fixes? Of course, I'd gladly be informed of any
    nonsense in my queries.
</pre>
 
Robert Sum <a href="mailto:sumrn@crd.ge.com"> replies: <p>
&lt;sumrn@crd.ge.com&gt;</a>
 
<pre>
    A two part answer:
 
      Part 1.  Neither version of mSQL has a particularly aggressive
    optimiser and they default to following the FROM and WHERE lines in
    the order in which they are entered while performing the query.
    Joins generally result in the formation of temporary tables (on the
    disk); consequently, they are highly sensitive to the amount of
    data processed.  If the order of the query is such that the
    temporary tables are large (i.e., the first clauses to be processed
    don't reduce the overall amount of data by much), then it will be
    slower.  Therefore, if one knows what tables and clauses will
    result in the smallest sizes for intermediate results, a query
    written with those tables and clauses considered first will run
    faster than other possible queries.  Though not apparent in these
    examples, some versions of mSQL 2.0 Bx also have a glitch in the
    query mechanism that causes extra parenthesis to generate extra
    temporary tables.  So, it is best to use parentheses only when
    necessary.  (mSQL 1.x does not do parentheses.)  Finally, there is
    an initialisation overhead whenever a table must be read/mapped
    into memory which will often cause the first time a query is
    executed to be slower than succeeding times, provided that
    preceding queries have not used the same tables.
 
      Part 2.  Each of mSQL 1.x and 2.0 Bx have various bugs in their
    LIKE operations: one of them being the ability to run off the end
    of the data into mismatch and core-dump lands.  While trying not to
    blow my own horn, the mailing list archives have my patches that
    fix most the problems.  However, the use of a bug-free LIKE in
    these specific examples would only be a performance penalty
    because  &quot;LIKE 'SUSI'&quot; is the same as &quot;= 'SUSI'&quot;.
</pre>
 
<hr>
 
<h3><a name="multi-table-selects">Why don't my SELECT statements work
across multiple tables?</a></h3>
 
Jonathan Loh <a href="mailto:jloh@netjet.com">
&lt;jloh@netjet.com&gt;</a> writes: <p>
 
<pre>
    I'm trying to do a join and each time it gives me an error.
 
	select * from t1,t2 where t1.key=t2.key
 
    This always gives me:
 
	Reference to un-selected table &quot;&quot;
 
    I'm using the 1.0.16 version
</pre>
 
Chris Lambrou <a href="mailto:chrislambrou@internetmci.com">
&lt;chrislambrou@internetmci.com&gt;</a> replies: <p>
 
<pre>
    There's one I've never seen! Try this instead:
 
	select t1.*, t2.* from t1,t2 where t1.key=t2.key
</pre>
 
And Joshua Kugler <a href="mailto:jkugler@inreach.com">
&lt;jkugler@inreach.com&gt;</a> also replies: <p>
 
<pre>
    I found it in the UMBC mSQL Tutorial, at
 
	<a href="http://www.cs.umbc.edu/~finin/461/msql">http://www.cs.umbc.edu/~finin/461/msql</a>:
 
    Minor limitation: can't use &quot;SELECT *&quot; over multiple tables
 
    In mSQL's SELECT you can not use * for the fields you want in the
    answer if there are two or more tables involved. Note that this
    limitation does not reduce the &quot;expressive power&quot; of mSQL
    but it is somewhat annoying. Here is an example. Suppose we define
    T1 and T2 as:
 
      create table t1 (a int, b int)
      create table t2 (b int, c int)
      insert into t1 values (1,2)
      insert into t2 values (2,3)
 
    Then these queries generate errors:
 
      mSQL &gt; select * from t1, t2 where t1.b=t2.b \g
      ERROR : Reference to un-selected table &quot;&quot;
 
      mSQL &gt; select * from t1,t2 \g
      ERROR : Reference to un-selected table &quot;&quot;
 
    But explicitly specifying the fields to return works:
 
      mSQL &gt; select t1.a,t1.b,t2.b,t2.c from t1, t2 \g
 
       +----------+----------+----------+----------+
       | a        | b        | b        | c        |
       +----------+----------+----------+----------+
       | 1        | 2        | 2        | 3        |
       +----------+----------+----------+----------+
 
      mSQL &gt; select t1.a,t1.b,t2.c from t1, t2 where t1.b=t2.b \g
 
       +----------+----------+----------+
       | a        | b        | c        |
       +----------+----------+----------+
       | 1        | 2        | 3        |
       +----------+----------+----------+
</pre>
 
<hr>
<hr width="75%">
 
<center>
<h2><a name="how-do-i">How do I ...?</a></h2>
</center>
 
<h3><a name="single-quotes">How do I embed single quotes in a
field?</a></h3>
 
To handle a single quote, escape it with a backslash character. So the
string <p>
 
<pre>
    'Bambi's'
</pre>
 
would be entered as <p>
 
<pre>
    'Bambi\'s'.
</pre>
 
<b>Note:</b> This applies when using
<a href="#support-progs"><tt>msql</tt></a> - the database monitor
program. If you're developing your own application you may have to
escape other characters that are special to the language you're using -
for example perl, C or tcl. <p>
 
<hr>
 
<h3><a name="escape-chars">What other characters need special
treatment?</a></h3>
 
When specifying table or field names or when inserting values into
fields, the only other character that requires special handling is the
mSQL escape character backslash. To handle a backslash, escape it with
another backslash. So the string <p>
 
<pre>
    c:\windows\system\
</pre>
 
would be entered as <p>
 
<pre>
    'c:\\windows\\system\\'
</pre>
 
When using regular expressions in queries of the form: <p>
 
<pre>
    SELECT table.column FROM table WHERE table.column LIKE 'regexp'
</pre>
 
the following rules apply: <p>
 
<center>
<table border=3 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<th colspan=6>
	    Use this table for mSQL versions 1.0.13 and below
	</th>
    </tr>
 
    <tr>
	<th>
	    To search for<br>
	    this character
	</th>
 
	<th>
	    Use this<br>
	    string
	</th>
 
	<th>
	    or this<br>
	    string
	</th>
 
	<th>
	    To search for<br>
	    this character
	</th>
 
	<th>
	    Use this<br>
	    string
	</th>
 
	<th>
	    or this<br>
	    string
	</th>
    </tr>
 
    <tr>
	<td align=center>
	    $
	</td>
 
	<td align=center>
	    \\\\$
	</td>
 
	<td align=center>
	    [$]
	</td>
 
	<td align=center>
	    [
	</td>
 
	<td align=center>
	    \\\\[
	</td>
 
	<td align=center>
	    [[]
	</td>
    </tr>
 
    <tr>
	<td align=center>
	    %
	</td>
 
	<td align=center>
	    \\%
	</td>
 
	<td align=center>
	    [%]
	</td>
 
	<td align=center>
	    \
	</td>
 
	<td align=center>
	    \\\\\\\\
	</td>
 
	<td align=center>
	    [\\\\]
	</td>
    </tr>
 
    <tr>
	<td align=center>
	    '
	</td>
 
	<td align=center colspan=2>
	    \'
	</td>
 
	<td align=center>
	    ^
	</td>
 
	<td align=center>
	    \\\\^
	</td>
 
	<td align=center>
	    [^]
	</td>
    </tr>
 
    <tr>
	<td align=center>
	    (
	</td>
 
	<td align=center>
	    \\\\(
	</td>
 
	<td align=center>
	    [(]
	</td>
 
	<td align=center>
	    _
	</td>
 
	<td align=center colspan=2>
	    \\_
	</td>
    </tr>
 
    <tr>
	<td align=center>
	    )
	</td>
 
	<td align=center>
	    \\\\)
	</td>
 
	<td align=center>
	    [)]
	</td>
 
	<td align=center>
	    |
	</td>
 
	<td align=center>
	    \\\\|
	</td>
 
	<td align=center>
	    [|]
	</td>
    </tr>
 
    <tr>
	<td align=center>
	    ?
	</td>
 
	<td align=center>
	    \\\\?
	</td>
 
	<td align=center>
	    [?]
	</td>
    </tr>
</table>
</center>
<p>
 
<center>
<table border=3 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<th colspan=4>
	    Use this table for mSQL versions 1.0.14 and above
	</th>
    </tr>
 
    <tr>
	<th>
	    To search for<br>
	    this character
	</th>
 
	<th>
	    Use this<br>
	    string
	</th>
 
	<th>
	    To search for<br>
	    this character
	</th>
 
	<th>
	    Use this<br>
	    string
	</th>
 
    <tr>
	<td align=center>
	    $
	</td>
 
	<td align=center>
	    [$]
	</td>
 
	<td align=center>
	    [
	</td>
 
	<td align=center>
	    [[]
	</td>
    </tr>
 
    <tr>
	<td align=center>
	    %
	</td>
 
	<td align=center>
	    [\\%]
	</td>
 
	<td align=center>
	    \
	</td>
 
	<td align=center>
	    <i>cannot be searched</i>
	</td>
    </tr>
 
    <tr>
	<td align=center>
	    '
	</td>
 
	<td align=center>
	    \'
	</td>
 
	<td align=center>
	    ^
	</td>
 
	<td align=center>
	    <i>cannot be searched</i>
	</td>
    </tr>
 
    <tr>
	<td align=center>
	    (
	</td>
 
	<td align=center>
	    [(]
	</td>
 
	<td align=center>
	    _
	</td>
 
	<td align=center>
	    \\_
	</td>
    </tr>
 
    <tr>
	<td align=center>
	    )
	</td>
 
	<td align=center>
	    [)]
	</td>
 
	<td align=center>
	    |
	</td>
 
	<td align=center>
	    [|]
	</td>
    </tr>
 
    <tr>
	<td align=center>
	    ?
	</td>
 
	<td align=center>
	    [?]
	</td>
 
	<td align=center>
	    ]
	</td>
 
	<td align=center>
	    []]
	</td>
    </tr>
</table>
</center>
<p>
 
<b>Note:</b> This applies when using
<a href="#support-progs"><tt>msql</tt></a> - the database monitor
program. If you're developing your own application you may have to
escape other characters that are special to the language you're using -
for example perl, C or tcl. <p>
 
If you are using mSQL via web interfaces, you may wish to use special
HTML characters instead. For example, a single quote can be
represented as: <p>
 
<pre>
    &amp;#39;
</pre>
 
A complete list of the HTML coded characters can be found at: <p>
 
<pre>
    <a href="http://www.w3.org/pub/WWW/MarkUp/html-spec/html-spec_13.html">http://www.w3.org/pub/WWW/MarkUp/html-spec/html-spec_13.html</a>
</pre>
 
While we're on the topic of regular expressions, this explanation from
Robert Sum <a href="mailto:sumrn@crd.ge.com">
&lt;sumrn@crd.ge.com&gt;</a> may shed some light on what
happens inside mSQL: <p>
 
Robert writes: <p>
 
<pre>
    In mSQL 1.0.x, any time you use LIKE, you use an unusual form of
    &quot;regular expressions&quot; which are the way they are because of some
    historical implementation decisions that, in retrospect, were,
    perhaps, not the right ones. Consequently, mSQL 1.0.x does the
    following for a LIKE pattern which is somewhere between globbing
    and full regular expressions:
 
    1. it tries to force a match of the whole data string by implicitly
       prepending the pattern with the beginning-of-pattern operator ^
       and appending the end-of-pattern operator $ to the pattern
       (which makes the use/non-use of these operators a little quirky
       [see below]),
    2. it does NOT allow the ., *, and + regexps (they are plain characters),
    3. it does allow character classes using [] (i.e., any single
       character within the [] is matched unless the first char is ^ in
       which case any single character not in the [] is matched),
    4. it does allow alternation using |,
    5. it does allow grouping using (),
    6. it does allow optional characters or groups using ?,
    7. it allows _ to represent any single character (what would
       ordinarily be the . above),
    8. it allows % to represent any string of characters (what would
       ordinarily be the combination .*),
 
    As you can derive from above, the characters
 
	^, $, [, ], |, (, ), ?, _, %, \
 
    are all special in one way or another. There is a way to search for
    everything, it just might not be completely obvious. For instance,
    using the msql program, one can use
 
       \\_ to match _
       \\% to match %
       \\^ to match ^
       \\$ to match $
       \\\\ to match \
 
    Basically, there are three places where \ is used as the escape
    character:
 
      1. The msql program (always treats \ as an escape character),
      2. The translation process (treats \ as an escape only when
	 followed by % or _), and
      3. The regular expression evaluator (always treats \ as an escape
         character).
 
    In the first two examples above, steps 1 and 2 strip backslashes.
    In the next three examples, steps 1 and 3 strip backslashes.
    Things are actually a bit more consistent than folks realize, I
    think. (Note:  If you are using Perl or Tcl or some such, then
    replace 1. with appropriate activity of that application.)
 
    Furthermore, a recent message asked about
 
	SELECT * FROM doc_info
	WHERE foo LIKE 'x' or foo LIKE 'y' or foo LIKE 'z'
	and bar LIKE 'a' or bar LIKE 'b' or bar LIKE 'c'
 
    to which I replied, try
 
	SELECT * FROM doc_info
	WHERE foo LIKE '(x)|(y)|(z)'
	AND bar LIKE '(a)|(b)|(c)'
 
    Well, I forgot about rule number 1. above, which means this
    conversion may not be quite right and the original sender may need
    to fudge things a bit using either an additional set of parenthesis
    to force a match of the whole data string as in
 
	SELECT * FROM doc_info
	WHERE foo LIKE '((x)|(y)|(z))'
	AND bar LIKE '((a)|(b)|(c))'
 
    or if he really wants a match anywhere within the string
 
	SELECT * FROM doc_info
	WHERE foo LIKE '$|(x)|(y)|(z)|^'
	AND foo &lt;&gt; ''
	AND bar LIKE '$|(a)|(b)|(c)|^'
	AND bar &lt;&gt; ''
 
    then this should work, but as noted, one must be careful about the
    empty string. (If you want the empty string, just leave out the
    not equal clauses.)
</pre>
 
<hr>
 
<h3><a name="null-fields">How do I handle null fields?</a></h3>
 
For the NULL values, just use the NULL keyword. <p>
 
For example <p>
 
<pre>
    insert into foo values ( NULL, 1, 2, 'some text' )
</pre>
 
<hr>
 
<h3><a name="no-case">How do I perform case insensitive
matches?</a></h3>
 
mSQL uses Henry Spencer's regular expression library with a few
modifications. To perform case insensitive searches, your query should
take the form <p>
 
<pre>
    SELECT * FROM my_table WHERE my_field LIKE '[Ss][Oo][Mm][Ee]
    [Vv][Aa][Ll][Uu][Ee]'
</pre>
 
Sol Katz's <a href="mailto:skatz@blm.gov">&lt;skatz@blm.gov&gt;</a>
<a href="#oo-html">Object Oriented HTML API</a> includes a C routine
that converts a string into its case insensitive form. You may wish to
use this in any C code that you write. See the section below on
<a href="#other-apps">&quot;Contributed Code and Third Party
Applications&quot;</a> <p>
 
Alternatively, you can create an additional field in each table that
will hold a single case version of the information you are likely to be
searching for. <p>
 
For perl users, Michael Cowden
<a href="mailto:cowden@leithaus.leitess.com">
&lt;cowden@leithaus.leitess.com&gt</a> has contributed this code
example: <p>
 
<pre>
    The following statement turns mSQL into [mM][sS][qQ][lL]
 
    $string = &quot;mSQL&quot;;
    $string =~ s/(.)/\[\L$1\E\U$1\E\]/g;
</pre>
 
Vivek Khera
<a href="mailto:khera@kci.kciLink.com">&lt;khera@kci.kciLink.com&gt;</a>
suggests a simpler method for perl users: <p>
 
<pre>
    Personally, I use this in Perl, as there is no need to complicate
    the regular expression with non-alpha characters.
 
       $string =~ s/([A-Za-z])/\[\L$1\U$1\]/gi;
</pre>
 
Rasmus Lerdorf's
<a href="mailto:rasmus@vex.net">&lt;rasmus@vex.net&gt;</a>
<a href="#php">Personal Home Page Construction Kit</a> includes built
in operations for case insensitive searches by way of its
<tt>msql_RegCase(string)</tt> command. <p>
 
Version 2 of mSQL will support functions similar to <tt>upper()</tt>
and <tt>lower()</tt> which will obviate the need for the above. <p>
 
<hr>
 
<h3><a name="add-column">How do I add a column to an existing
table?</a></h3>
 
You can't! Once a table is created it must stay the way it is. <p>
 
One possible solution is to use
<a href="#support-progs"><tt>msqldump</tt></a> to create an ASCII dump
of the entire database. Then edit this dump file by hand and add the
extra field to the CREATE clause. You'll also need to edit each INSERT
clause to ensure that the new field is referenced. Once you've
modified the dump file, drop and recreate the database using
<a href="#support-progs"><tt>msqladmin</tt></a> and re-populate the new
database using the dump file and
<a href="#support-progs"><tt>msql</tt></a>. <p>
 
This procedure could be automated by a shell or perl script. <p>
 
<b>Note:</b> Use the <tt>-c</tt> option to
<a href="#support-progs"><tt>msqldump</tt></a> to ensure that a
complete dump of the table is produced. <p>
 
As an example consider this output from
<a href="#support-progs"><tt>msqldump</tt></a>
 
<pre>
    #
    # mSQL Dump  (requires mSQL-1.0.6 or better)
    #
    # Host: localhost    Database: test
    #--------------------------------------------------------
 
 
    #
    # Table structure for table 'test'
    #
    CREATE TABLE test (
      name CHAR(40),
      num INT
    ) \g
 
 
    #
    # Dumping data for table 'test'
    #
 
    INSERT INTO test VALUES ('item 999',999)\g
    ...
    INSERT INTO test VALUES ('item 0',0)\g
</pre>
 
If you wish to insert a field, say &quot;discount&quot;, then you will
need to modify the dump file as follows: <p>
 
<pre>
    #
    # mSQL Dump  (requires mSQL-1.0.6 or better)
    #
    # Host: localhost    Database: test
    #--------------------------------------------------------
 
 
    #
    # Table structure for table 'test'
    #
    CREATE TABLE test (
      name CHAR(40),
      num INT,
      discount REAL
    ) \g
 
 
    #
    # Dumping data for table 'test'
    #
 
    INSERT INTO test VALUES ('item 999',999,0.0)\g
    ...
    INSERT INTO test VALUES ('item 0',0,0.0)\g
</pre>
 
<b>Notice</b> that every insert clause MUST be changed as well as the
table definition. <p>
 
<hr>
 
<h3><a name="parent-child">When should I call <tt>msqlConnect()</tt> in
a parent/child situation?</a></h3>
 
If both the parent and child processes want to talk to the mSQL server
then you must call <tt>msqlConnect()</tt> <b>AFTER</b> you fork. They
mustn't share the same socket. <p>
 
<hr>
 
<h3><a name="reserved-words">Can I use mSQL reserved words as field or
table names?</a></h3>
 
No. The mSQL parser gets very confused if you attempt to use reserved
words as the name of a table or field. The full list of reserved words
for both version 1.<i>x</i> and 2.<i>x</i> (as obtained from
<tt>msql_lex.c</tt>) is: <p>
 
<center>
<table border=2 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<th colspan=8>
	    Reserved words in mSQL 1.<i>x</i>
	</th>
    </tr>
 
    <tr>
	<td>&lt;</td>
	<td>&gt;=</td>
	<td>by</td>
	<td>distinct</td>
	<td>integer</td>
	<td>not</td>
	<td>real</td>
	<td>update</td>
    </tr>
 
    <tr>
	<td>&lt;=</td>
	<td>all</td>
	<td>char</td>
	<td>drop</td>
	<td>into</td>
	<td>null</td>
	<td>select</td>
	<td>values</td>
    </tr>
 
    <tr>
	<td>&lt;&gt;</td>
	<td>and</td>
	<td>create</td>
	<td>from</td>
	<td>key</td>
	<td>or</td>
	<td>set</td>
	<td>where</td>
    </tr>
 
    <tr>
	<td>=</td>
	<td>as</td>
	<td>delete</td>
	<td>insert</td>
	<td>like</td>
	<td>order</td>
	<td>smallint</td>
    </tr>
 
    <tr>
	<td>&gt;</td>
	<td>asc</td>
	<td>desc</td>
	<td>int</td>
	<td>limit</td>
	<td>primary</td>
	<td>table</td>
    </tr>
</table>
</center>
<p>
 
<center>
<table border=2 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<th colspan=10>
	    Reserved words in mSQL 2.0B6
	</th>
    </tr>
 
    <tr>
	<td>
	    &lt;
	</td>
 
	<td>
	    all
	</td>
 
	<td>
	    by
	</td>
 
	<td>
	    delete
	</td>
 
	<td>
	    insert
	</td>
 
	<td>
	    limit
	</td>
 
	<td>
	    on
	</td>
 
	<td>
	    select
	</td>
 
	<td>
	    sum
	</td>
 
	<td>
	    update
	</td>
    </tr>
 
    <tr>
	<td>
	    &lt;=
	</td>
 
	<td>
	    and
	</td>
 
	<td>
	    char
	</td>
 
	<td>
	    desc
	</td>
 
	<td>
	    int
	</td>
 
	<td>
	    max
	</td>
 
	<td>
	    or
	</td>
 
	<td>
	    sequence
	</td>
 
	<td>
	    table
	</td>
 
	<td>
	    value
	</td>
    </tr>
 
    <tr>
	<td>
	    &lt;&gt;
	</td>
 
	<td>
	    as
	</td>
 
	<td>
	    clike
	</td>
 
	<td>
	    distinct
	</td>
 
	<td>
	    integer
	</td>
 
	<td>
	    min
	</td>
 
	<td>
	    order
	</td>
 
	<td>
	    set
	</td>
 
	<td>
	    text
	</td>
 
	<td>
	    values
	</td>
    </tr>
 
    <tr>
	<td>
	    =
	</td>
 
	<td>
	    asc
	</td>
 
	<td>
	    count
	</td>
 
	<td>
	    drop
	</td>
 
	<td>
	    into
	</td>
 
	<td>
	    money
	</td>
 
	<td>
	    primary
	</td>
 
	<td>
	    slike
	</td>
 
	<td>
	    time
	</td>
 
	<td>
	    where
	</td>
    </tr>
 
    <tr>
	<td>
	    &gt;
	</td>
 
	<td>
	    avg
	</td>
 
	<td>
	    create
	</td>
 
	<td>
	    from
	</td>
 
	<td>
	    key
	</td>
 
	<td>
	    not
	</td>
 
	<td>
	    real
	</td>
 
	<td>
	    smallint
	</td>
 
	<td>
	    uint
	</td>
 
    </tr>
 
    <tr>
	<td>
	    &gt;=
	</td>
 
	<td>
	    avl
	</td>
 
	<td>
	    date
	</td>
 
	<td>
	    index
	</td>
 
	<td>
	    like
	</td>
 
	<td>
	    null
	</td>
 
	<td>
	    rlike
	</td>
 
	<td>
	    step
	</td>
 
	<td>
	    unique
	</td>
 
    </tr>
</table>
</center>
<p>
 
Remember that mSQL reserved words are case insensitive so UPPER case or
MiXeD cAsE reserved words are also forbidden in table or field names.
<p>
 
<hr>
 
<h3><a name="maxmin">How do I find the maximum or minimum value in a
table?</a></h3>
 
To obtain the maximum value use: <p>
 
<pre>
    SELECT number FROM table ORDER BY number DESC LIMIT 1
</pre>
 
To obtain the minimum value use: <p>
 
<pre>
    SELECT number FROM table ORDER BY number LIMIT 1
</pre>
 
This will only work with mSQL 1.0.9 and above unless you have applied
the unofficial LIMIT patch to earlier versions. See the
<a href="#list-archives">mSQL mailing list archives</a> for details on
this patch. (Before searching for this unofficial patch, you should
seriously consider upgrading to the
<a href="#current-version">latest</a> version of mSQL). <p>
 
<b>Note:</b> Rasmus Lerdorf writes: <p>
 
<pre>
    The LIMIT statement limits the number of records actually
    transferred from the server to the client. It doesn't limit the
    scope of the search at all in any way. That means that if you are
    looking for the maximum value in a table with 30,000 entries, the
    query will first build the entire sorted result in memory in the
    server, but when it comes time to transferring the result to the
    client, it only sends the first item.
 
    In many cases, especially when you have a lot of fields, or long
    fields, the time it takes to transfer the data from the server to
    the client is actually many times that of the actual search. And
    the msqld daemon is tied up and not available to other clients
    while it is wasting time sending result records that will never be
    used. So, if you do queries and you know you will only be looking
    at the first couple of them, you should use the limit clause and
    cut down on the amount of useless records being sent across the
    socket.
</pre>
 
<hr>
 
<h3><a name="schema">How can I determine the structure of a database?</a></h3>
 
Use the <a href="#support-progs"><tt>relshow</tt></a> application that
comes bundled with the mSQL software distribution. <p>
 
<ul>
    <li> To discover which databases are available use: <p>
 
<pre>
    relshow
</pre>
 
    or <p>
 
<pre>
    relshow -h <i>host</i>
</pre>
 
    This will return output similar to: <p>
 
<pre>
	+-----------------+
	|    Databases    |
	+-----------------+
	| test            |
	| photos          |
	| patches         |
	+-----------------+
</pre>
 
    <li> To discover which tables are contained within a database use: <p>
 
<pre>
    relshow <i>dbname</i>
</pre>
 
    or <p>
 
<pre>
    relshow -h <i>host</i> <i>dbname</i>
</pre>
 
    This will return output similar to: <p>
 
<pre>
 
    Database = test
 
	+---------------------+
	|       Table         |
	+---------------------+
	| test_table          |
	| addresses           |
	| telephone           |
	+---------------------+
</pre>
 
    <li> To discover the structure of a particular table use: <p>
 
<pre>
    relshow <i>dbname</i> <i>tablename</i>
</pre>
 
    or <p>
 
<pre>
    relshow -h <i>host</i> <i>dbname</i> <i>tablename</i>
</pre>
 
    This will return output similar to: <p>
 
<pre>
    Database = test
 
    Table    = test_table
 
	+-----------------+----------+--------+----------+-----+
	|     Field       |   Type   | Length | Not Null | Key |
	+-----------------+----------+--------+----------+-----+
	| name            | char     | 40     | N        | N   |
	| num             | int      | 4      | N        | N   |
	+-----------------+----------+--------+----------+-----+
</pre>
 
</ul>
 
<hr>
 
<h3><a name="reconnect">What happens when the mSQL server goes down
between requests?</a></h3>
 
If the mSQL database server process,
<a href="#support-progs"><tt>msqld</tt></a>, dies and is subsequently
restarted, or the host on which it was running is rebooted, any
processes that were connected to the mSQL database server MUST be
reconnected. This is not unique to mSQL, the Oracle database server
behaves in a similar manner. <p>
 
Programs that were connected to the mSQL database server should be
either restarted or have some internal mechanism whereby they notice
the server has died and attempt a reconnection. <p>
 
One possible method for checking the status of the database server
would be to examine the return status of the
<a href="#api"><tt>msqlSelectDB()</tt></a> call. <p>
 
<hr>
 
<h3><a name="muli-servers">Can I run more than one copy of
<tt>msqld</tt> on the same CPU?</a></h3>
 
Steve A. Olson writes: <p>
 
<pre>
    I'm looking for a way to provide the full 25 connections to each of
    many mSQL databases running on a single box. Here's an idea, will
    it work? or is there a better way?
 
    (korn shell example)
    $ export MSQL_TCP_PORT=3000; msqld
    $ export MSQL_TCP_PORT=3001; msqld
    $ export MSQL_TCP_PORT=3002; msqld
 
    Then connect to the database as follows:
    $ export MSQL_TCP_PORT=3000; msql db_a
 
    While the above runs, another user connects:
    $ export MSQL_TCP_PORT=3001; msql db_b
</pre>
 
David Hughes replies: <p>
 
<pre>
    Well, sort of. By running 'msql db_a' you are using the local UNIX
    socket not the TCP socket so you'd have to use MSQL_UNIX_PORT not
    MSQL_TCP_PORT.
 
    The other thing is that you should run 3 MSQL_HOME areas
    (/usr/local/Minerva for example). If two of these servers __ever__
    access the same database at the same time then you are stuffed.
 
    So, something like
 
    export MSQL_UNIX_PORT=/dev/msql_1; export MSQL_HOME=/Minerva1; msqld&amp;
    export MSQL_UNIX_PORT=/dev/msql_2; export MSQL_HOME=/Minerva2; msqld&amp;
 
    and
 
    export MSQL_UNIX_PORT=/dev/msql_1; msql db_a
    export MSQL_UNIX_PORT=/dev/msql_2; msql db_b
 
    would do the job.
</pre>
 
<hr>
 
<h3><a name="unique-keys">How can I automatically ensure that each
record receives a unique primary key?</a></h3>
 
mSQL version 1.<i>x</i> does not have an automatic key assignment
functionality. A number of solutions to this problem are available. <p>
 
<ul>
    <li> Use Pascal Forget's <a href="#sgs">unique sequence number
    generator</a>. <p>
 
    <li> Build the unique key assignment and management into your
    application. <p>
 
    One suggestion on how to implement this comes from Vivek Khera
    <a href="mailto:khera@kci.kciLink.com">
    &lt;khera@kci.kciLink.com&gt;</a>. <p>
 
    Vivek writes: <p>
 
<pre>
    What I do is take some of the fields in the record, tack on a salt
    like the current time and generate a hash (either SHA or MD5) of
    it. I use part of the the hash string value as the key.
</pre>
 
    Another solution is provided by Rasmus Lerdorf
    <a href="mailto:rasmus@vex.net">&lt;rasmus@vex.net&gt;</a>. <p>
 
    Rasmus writes: <p>
 
<pre>
    The issue here is not so much how to generate a key, but how to
    ensure it is unique. The way I have done it in the past is to
    associate a lock file with each table. Each table has a counter
    record. With the table locked, I pull out the current counter
    value, increment it, and put it back. Then I unlock the table. This
    is not the most efficient way to do it, but it does work nicely for
    systems that do not get pounded with queries.
</pre>
 
    Others have suggested using a timestamp with milli second
    granularity. This approach has its pitfalls. <p>
 
    <ul>
	<li> Not all client side operating systems have access to the
	<tt>gettimeofday()</tt> calls (or its equivalent).
 
	<li> Not all systems can determine the system time to the milli
	second.
 
	<li> Because the timestamps are being determined by the
	clients, it is possible for two clients to arrive at the same
	timestamp. This problem is amplified if you have a number of
	clients running on different machines which are not
	synchronised. This can be avoided by recalculating the
	timestamp if an insert operation fails due to a duplicate key
	error.
 
	<li> Clocks may drift backwards causing old timestamps to be
	reused. Sites that use time synchronisation methods such as
	<tt>rdate</tt> or <tt>xntp</tt> may be prone to this.
    </ul>
    <p>
 
    <li> Use a patch provided Jochem Wiedmann
    <a href="mailto:wiedmann@wiedmann.neckar-alb.de">
    &lt;wiedmann@wiedmann.neckar-alb.de&gt;</a>. This patch allows you
    to declare a key as: <p>
 
<pre>
    AUTO PRIMARY KEY
</pre>
 
    This patch is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLpatches/msql-auto-pkey.patch.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLpatches/msql-auto-pkey.patch.gz</a>
    (6329 bytes)
</pre>
 
    <li> Marcantonio Magnarapa
    <a href="mailto:marcantonio@magnarapa.com">
    &lt;marcantonio@magnarapa.com&gt;</a> submits this suggestion for
    those of you with web interfaces to your mSQL databases: <p>
 
    Marcantonio writes:
 
<pre>
    I have found a quick and dirty method that works for most
    situations. In my case, I include in the web form where I insert
    something into the database a HIDDEN field whose value is
    determined by a webpage hit counter. Any text-based server-side
    include hit counter is fine. For instance, in my page I have
    something like:
 
    &lt;input type=hidden name=id value=&quot;&lt;!--#exec cgi-bin/counter.pl&gt;&quot; &gt;
 
    the ID field is always different, although it may not be sequential.
    this didn't require a single line of code because I already had a
    hitcounter, and this is supposed to work for many situations.
</pre>
 
</ul>
 
mSQL version 2.<i>x</i> has an automatic sequence number associated
with each table. More on this in later FAQs. <p>
 
<hr>
 
<h2><a name="multiple-includes">How can I avoid compiler redefinition
errors when compiling my own mSQL applications?</a></h2>
 
When building your own mSQL applications you may encounter compiler
errors similar to: <p>
 
<pre>
    /usr/local/Minerva/include/msql.h:30: redefinition of `m_row'
    /usr/local/Minerva/include/msql.h:32: redefinition of `struct field_s'
</pre>
 
This occurs because the mSQL header file <tt>msql.h</tt> has been
included more than once. <p>
 
To avoid this, apply the following patch to <tt>msql.h</tt>
contributed by Vesa Tuomi <a href="mailto:vesa@cardinal.fi">
&lt;vesa@cardinal.fi&gt;</a> <p>
 
<pre>
<font size="-1">
*** ./src/msql/msql.h.orig Wed Mar  6 09:27:20 1996
--- ./src/msql/msql.h      Thu Mar  7 10:29:46 1996
***************
*** 16,21 ****
--- 16,23 ----
  **
  */
 
+ #ifndef       __MSQL_H__
+ #define       __MSQL_H__
 
  #if defined(__STDC__) || defined(__cplusplus)
  #  define __ANSI_PROTO(x)     x
***************
*** 109,111 ****
--- 111,115 ----
  #ifdef __cplusplus
        }
  #endif
+
+ #endif        /* __MSQL_H__ */
</font>
</pre>
 
and rerun the <tt>make install</tt> phase of the mSQL installation
procedure. This will remake all the core mSQL applications and install
the modified <tt>msql.h</tt> file in your installation <tt>include</tt>
directory. It will also reinstall the core mSQL applications in your
installation <tt>bin</tt> directory. There is NO need to recompile any
other third party applications. <p>
 
<hr>
 
<h3><a name="linking">How do I link the mSQL library with my own
code?</a></h3>
 
Most compilers will search a well defined list of standard directories
for include and library files. These are typically
<tt>/usr/include</tt>, <tt>/usr/lib</tt>, <tt>/lib</tt> and
occasionally <tt>/usr/local/include</tt> and <tt>/usr/local/lib</tt>.
If you wish to use files that are outside these locations you must tell
your compiler which directories to search. <p>
 
Most C compilers understand the command line arguments: <p>
 
<pre>
    -I<i>include_directory</i>
</pre>
 
and
 
<pre>
    -L<i>library_directory</i>
</pre>
 
to mean &quot;search the directory <i>include_directory</i> for
include files and search the directory <i>library_directory</i> for
libraries&quot;. <p>
 
If you wish to compile a program that will communicate with the mSQL
database server you must tell the compiler where the mSQL include and
library files were installed. Assuming you installed mSQL in
<tt>/usr/local/Minerva</tt> and you use <tt>gcc</tt>, the syntax would
resemble: <p>
 
<pre>
    gcc -c -I/usr/local/Minerva/include <i>your_prog.c</i>
    gcc -o <i>your_prog</i> <i>your_prog.o</i> -L/usr/local/Minerva -lmsql
</pre>
 
<b>Note:</b> Libraries <b>MUST</b> be specified last. Most compilers
will give strange errors if you specify the library paths and libraries
before any <tt>.c</tt> or <tt>.o</tt> files. <p>
 
<b>Note:</b> that the mSQL library name has been truncated. The actual
mSQL library file name is <tt>libmsql.a</tt>. However, the compiler
(and the link loader) only need the unique part of the file name so
the <tt>lib</tt> and <tt>.a</tt> components should be removed when
passing the library name on the command line. <p>
 
If you are using an SVR4 version of Unix (such as Solaris 2.<i>x</i>)
you may also have to include some networking libraries: <p>
 
<pre>
    gcc -o <i>your_prog</i> <i>your_prog.o</i> -lsocket -lnsl -L/usr/local/Minerva -lmsql
</pre>
 
<hr>
 
<h3><a name="numrows">How can I find the number of rows in a
table?</a></h3>
 
If you're writing code in C, the following will work: <p>
 
<pre>
    msqlQuery(dbsocket, &quot;SELECT * from blah&quot;);
    result = msqlStoreResult();
    number = msqlNumRows(result);
    msqlFreeResult(result);
</pre>
 
<hr>
 
<h3><a name="delete-all">How can I delete all the rows in a table
without destroying the table structure?</a></h3>
 
If you wish to clear the contents of a table without actually dropping
the table, use the following syntax: <p>
 
<pre>
    DELETE FROM <i>table_name</i>\g
</pre>
 
<hr>
 
<h3><a name="ascii-import">How can I import flat files or other
database formats into an mSQL table?</a></h3>
 
For users of mSQL version 1.<i>x</i>, use Pascal Forget's
<a href="mailto:pascal@wsc.com">&lt;pascal@wsc.com&gt;</a>
<tt>msql-import</tt> utility. Details on how to obtain this utility can
be found in the <a href="#other-apps"> &quot;Contributed Code and Third
Party Applications&quot;</a> section. Look for the subsection titled
<a href="#msql-import">&quot;Flat File Importer&quot;</a>. <p>
 
<tt>msql-import</tt> is bundled with mSQL version 2.<i>x</i>. <p>
 
<hr>
 
<h3><a name="touched">How can I determine the number of rows affected
by my last query?</a></h3>
 
David Hughes
<a href="mailto:bambi@hughes.com.au">
&lt;bambi@hughes.com.au&gt;</a> writes: <p>
 
<pre>
    In mSQL 2.0 the return value from msqlQuery() tells you the number
    of rows &quot;touched&quot; by the query (i.e. deleted for a
    DELETE, returned for a SELECT, modified for an UPDATE etc).
</pre>
 
<hr>
 
<h3><a name="w3auth">How can I get <tt>w3-msql</tt> to work with
<tt>w3-auth</tt>?</a></h3>
 
Jason Hudgins <a href="mailto:hudginsj@smtp.dancooks.com">
&lt;hudginsj@smtp.dancooks.com&gt;</a> has written a paper on
&quot;<i>The Unofficial Guide to W3-AUTH</i>&quot;. It is available
from: <p>
 
<pre>
    <a href="http://www.dancooks.com/~jason/w3-msql/w3auth.html">http://www.dancooks.com/~jason/w3-msql/w3auth.html</a>
</pre>
 
<hr>
<hr width="75%">
 
<!-- End of Part 3 -->
 
<center>
<h2><a name="other-apps">Contributed Code and Third Party
Applications</a></h2>
</center>
 
<b>Note:</b> Many of the URLs in this section have changed to reflect
a directory reorganisation at
<a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib">
<tt>ftp://bond.edu.au/pub/Minerva/msql/Contrib</tt></a>. The packages
in question will not be flagged as changed unless the package itself
has been updated since the last release of the FAQ. <p>
 
A number of people have contributed additional software that works with
mSQL. The contributed software falls into two categories - that
developed by David Hughes (mSQL's author) and that developed by
others. <p>
 
<b>Note:</b> Please consult the documentation that comes with each of
these applications to determine the licensing obligations that may be
involved in their use. <p>
 
<b>Note:</b> Some of these applications are available via anonymous ftp
from the mSQL contributed code directory
<a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib">
<tt>ftp://bond.edu.au/pub/Minerva/msql/Contrib</tt></a>. There is often
a delay of a day or so for newly released files to be moved from the
<a href="ftp://bond.edu.au/pub/Minerva/msql/Incoming">
<tt>ftp://bond.edu.au/pub/Minerva/msql/Incoming</tt></a> directory to
the <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib">
<tt>ftp://bond.edu.au/pub/Minerva/msql/Contrib</tt></a> directory. <p>
 
<center>
<table border=2 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<td>
	    <a href="#esl">ESL</a>
	</td>
 
	<td>
	    <a href="#msqltools">mSQL Tools</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#w3-msql">w3-msql</a>
	</td>
 
	<td>
	    <a href="#msqlui">mSQL User Interface for Windows</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#lite">Lite</a>
	</td>
 
	<td>
	    (=) <a href="#msutil">msutil - an alternative C API for mSQL</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#addfsecure">Addf-Secure</a>
	</td>
 
	<td>
	    <a href="#ms-windows">MS Windows</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#apache">Apache</a>
	</td>
 
	<td>
	    <a href="#nextstep">NeXTSTEP EOF</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#applixware">Applixware</a>
	</td>
 
	<td>
	    <a href="#oo-html">Object Oriented HTML API</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#backups">Backup Script</a>
	</td>
 
	<td>
	    <a href="#ocx">OCX</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#dns-bind">Bind</a>
	</td>
 
	<td>
	    <a href="#odbc">ODBC</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#command-line">Command Line Tool</a>
	</td>
 
	<td>
	    <a href="#onyx">Onyx</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#dbadmin">Dbadmin</a>
	</td>
 
	<td>
	    (=) <a href="#oscafe">Open Systems Cafe</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#dbase">DBASE</a>
	</td>
 
	<td>
	    <a href="#os2">OS/2</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#dbi">DBI/DBD</a>
	</td>
 
	<td>
	    <a href="#perl">Perl</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#dbunk">DBunk</a>
	</td>
 
	<td>
	    <a href="#php">PHP - Home Page Construction Kit</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#dbview">Dbview</a>
	</td>
 
	<td>
	    <a href="#pts">PTS</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#digger">Digger</a>
	</td>
 
	<td>
	    <a href="#python">Python</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#emacs">Emacs</a>
	</td>
 
	<td>
	    <a href="#radius">RADIUS</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#msql-import">Flat File Importer</a>
	</td>
 
	<td>
	    <a href="#rexx">REXX</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#html-example">HTML Interface to mSQL</a>
	</td>
 
	<td>
	    <a href="#simple-sql">Simple SQL</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#ici">ICI</a>
	</td>
 
	<td>
	    <a href="#sqlbase">SQLBase</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#jate">Jate</a>
	</td>
 
	<td>
	    <a href="#tcl">Tcl</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#java">Java</a>
	</td>
 
	<td>
	    <a href="#time-utils">Time and date utilities</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#jio">Jio</a>
	</td>
 
	<td>
	    <a href="#tkmsql">tkmSQL</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#siod">LISP/Scheme</a>
	</td>
 
	<td>
	    <a href="#sgs">Unique sequence number generator</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#literepository">Lite Repository</a>
	</td>
 
	<td>
	    <a href="#vdb">Virtual Database</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#metahtml">Meta-HTML</a>
	</td>
 
	<td>
	    <a href="#virtuflex">VirtuFlex</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#ms-access7">MS Access to mSQL</a>
	</td>
 
	<td>
	    <a href="#visual-basic">Visual Basic</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#ms-access-dbase3">MS Access and dBase III</a>
	</td>
 
	<td>
	    <a href="#wdb">WDB</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#msqlcgi">mSQL CGI</a>
	</td>
 
	<td>
	    <a href="#websql">Websql</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#msqlexpire">mSQLexpire</a>
	</td>
 
	<td>
	    <a href="#tryc-repository">Wojciech Tryc's Repository</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#msqlexport">mSQL Export</a>
	</td>
 
	<td>
	    <a href="#xfsql">XfSQL</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#msqlkeeper">mSQL Keeper</a>
	</td>
 
	<td>
	    <a href="#xsqlmenu">Xsqlmenu</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#msqlsummary">mSQL Summary</a>
	</td>
 
	<td>
	    <a href="#cplusplus">Z Classes for C++</a>
	</td>
    </tr>
</table>
</center>
<p>
 
<dl>
<dt><b><a name="esl">ESL</a></b>
 
    <dd> David developed mSQL as the database component of a larger
    network management project called Minerva (Minerva was the Roman
    goddess of knowledge and information). Another component of Minerva
    is an <b>E</b>xtensible <b>S</b>cripting <b>L</b>anguage called
    ESL. This has a C like syntax and provides support for the complete
    mSQL API as well as full SNMP support. It is available via
    anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/esl/esl-0.3.tar.gz">ftp://bond.edu.au/pub/Minerva/esl/esl-0.3.tar.gz</a>
    (407046 bytes)
</pre>
 
<dt><b><a name="w3-msql">W3-mSQL</a></b>
 
    <dd> David has also developed W3-mSQL which is an interface between
    the World-Wide Web and mSQL. It is a mechanism that can be used to
    greatly simplify the use of an SQL database behind a web server.
    Using W3-mSQL, you can embed SQL queries within your pages and have
    the results generated on the fly. It is available via anonymous ftp
    from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/w3-msql/w3-msql-1.0.3.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/w3-msql/w3-msql-1.0.3.tar.gz</a>
    (54811 bytes)
</pre>
 
    W3-mSQL version 2.0 is bundled with mSQL 2.<i>x</i>. It uses the
    bundled <a href="#lite">Lite</a> language embedded within HTML
    tags. <p>
 
    W3-mSQL version 2.0 is incompatible with version 1.<i>x</i>. <p>
 
    More details on version 2.0 of W3-mSQL can be found in the
    documentation distributed with mSQL 2.<i>x</i>. <p>
 
<dt><b><a name="lite">Lite</a></b>
 
    <dd> Bundled with mSQL version 2.<i>x</i> is Lite. Lite has a C
    like syntax and is designed as a stand alone scripting language for
    use with mSQL. <p>
 
    More details on Lite can be found in the documentation distributed
    with mSQL 2.<i>x</i>. <p>
 
<dt><b><a name="addfsecure">Addf-Secure</a></b>
 
    <dd> Max Levchin <a href="mailto:mlevchin@ampere.scale.uiuc.edu">
    &lt;mlevchin@ampere.scale.uiuc.edu&gt;</a> has written a utility
    that will allow you to securely add a new column to an mSQL table.
    More details can be found at: <p>
 
<pre>
    <a href="http://ampere.scale.uiuc.edu/~mlevchin/addf">http://ampere.scale.uiuc.edu/~mlevchin/addf</a>
</pre>
 
<dt><b><a name="apache">Apache</a></b>
 
    <dd> Dirk van Gulik
    <a href="mailto:Dirk.vanGulik@jrc.it">
    &lt;Dirk.vanGulik@jrc.it&gt;</a> has developed some code for use
    with HTML and HTTP servers with specific reference to the Apache
    HTTP server. It is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.ceo.org/pub/ewse-mSQL-apache-demos/apache-msql-demo.1.0.1.tar.gz">ftp://ftp.ceo.org/pub/ewse-mSQL-apache-demos/apache-msql-demo.1.0.1.tar.gz</a>
    (12723 bytes)
</pre>
 
    If you require a module that allows the Apache httpd daemon to
    perform authentication via an mSQL database, grab the file: <p>
 
<pre>
    <a href="http://www.apache.org/dist/contrib/modules/mod_auth_msql.c">http://www.apache.org/dist/contrib/modules/mod_auth_msql.c</a>
    (9269 bytes)
</pre>
 
<dt><b><a name="applixware">Applixware</a></b>
 
    <dd> Cameron Newham <a href="mailto:cam@sspl.demon.co.uk">
    &lt;cam@sspl.demon.co.uk&gt;</a> has written a shared library
    and query utility for Applixware, interfacing to MSQL 2.0. You can
    find more details at: <p>
 
<pre>
    <a href="http://www.iinet.com.au/~cam/applix.html">http://www.iinet.com.au/~cam/applix.html</a>
</pre>
 
<dt><b><a name="backups">Backup Script</a></b>
 
    <dd> C Latham <a href="mailto:clatham@nerosworld.com">
    &lt;clatham@nerosworld.com&gt;</a> has written a shell script that
    will backup mSQL databases. It is reprinted here in full: <p>
 
<pre>
<font size="-1">
#!/bin/sh
#       &#36;Id: msqlbckp v. 1.0 1996/6/22 &#36;
#       by: clatham@nerosworld.com
#
#       usage: msqlbckp [-h host] 'backup directory'
#
#   This script creates daily backups of mSQL databases. Input parameters are
#   host machine (if msqld is not running on local machine) and the pathname
#   of a directory where the backups will be made. The backups will be named:
#   {table name}.{day of week}.gz  (They are gzipped).
#
#   The structure of the backup files are in a format that can be read
#   by the 'msql' program to completely restore the database, by first
#   dropping the corrupt table, then recreating the table and populating it
#   with data (it is required that the database itself already exists, which
#   may require the database administrator to use msqladmin to create the
#   database in extreme circumstances).
#
#   Access is required to msqldump, relshow and gzip.
#
#   Set crontab to execute this script at, say, 4 am every day, every other
#   day, or however often you want backups of your databases.
#
 
err( ) {
	echo usage: msqlbckp [-h host] 'backup directory'
	exit 1
}
 
case $# in
1)
	bdir=$1
	;;
3)
	if [ $1 = &quot;-h&quot; ]
	then
		host=&quot;-h &quot;$2
	else
		err
	fi
	bdir=$3
	;;
*)
	err
	;;
esac
 
dow=`date '+%A'`   # Get the day of the week
 
# Use relshow to get a list of the available databases, and pare that down
# into a file listing one database name per line...
relshow ${host} | \
	sed -n -e '/^....[ \-].*/d' -e 's/  \| //' -e 's/ *\|//p' \
	&gt;${bdir}/db.names
 
# Get each database name from the file for table processing
dbline=1
while [ 1 ]     # Do forever until no more db names
do
	dodb=`cat ${bdir}/db.names | sed -n &quot;${dbline}p&quot;`
	if [ &quot;${dodb}&quot; = &quot;&quot; ]   # if no more db names to do
	then
		break
	fi
 
	cf=${bdir}/${dodb}.${dow}       # Define current working file
 
echo &quot;#
# mSQL Dump of Database: ${dodb}
#
# Begin by dropping all tables
#---------------------------------------------
 
&quot; &gt; ${cf}
 
	# Now add commands to 'drop' each table in the database
	relshow ${host} ${dodb} | \
	sed -n -e '/^....[ \-].*/d' -e 's/  \| //' -e 's/ *\|//p' | \
	awk '($0 !~ /^$/) { print &quot;DROP TABLE&quot;, $1, &quot;\\g&quot; }' &gt;&gt; ${cf}
 
	echo &quot; &quot; &gt;&gt; ${cf}
 
	# Dump the database structure and data into the backup file
	msqldump ${host} ${dodb} &gt;&gt; ${cf}
 
	# Finally, gzip the file
	gzip -f ${cf}
 
	# Next database name
	dbline=`expr ${dbline} + 1`
done
 
rm ${bdir}/db.names     # Get rid of temp file
</font>
</pre>
 
<dt><b><a name="dns-bind">Bind</a></b>
 
    <dd> Chris Seawood
    <a href="mailto:mgrcls@nextwork.rose-hulman.edu">
    &lt;mgrcls@nextwork.rose-hulman.edu&gt;</a> has extended the DNS
    naming service - BIND - to support mSQL databases. It is available
    in the <tt>contrib/msql</tt> directory of the current release of
    BIND which can be obtained via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.vix.com/pub/bind/release/bind-4.9.3-REL.tar.gz">ftp://ftp.vix.com/pub/bind/release/bind-4.9.3-REL.tar.gz</a>
    (1682741 bytes)
</pre>
 
    Chris has also released a beta version of <tt>msql_bind</tt> that
    works with mSQL version 2.<i>x</i>. For more details see: <p>
 
<pre>
    <a href="http://www.seawood.org/msql_bind">http://www.seawood.org/msql_bind</a>
</pre>
 
<dt><b><a name="command-line">Command Line Tool</a></b>
 
    <dd> Kai Mysliwiec
    <a href="mailto:kvm@camelot.de">&lt;kvm@camelot.de&gt;</a>
    has developed an mSQL tool that allows you to send SQL queries from
    the command line. It is available from: <p>
 
<pre>
    <a href="http://www.camelot.de/~kvm/progs/sql.tar.gz">http://www.camelot.de/~kvm/progs/sql.tar.gz</a>
    (14083 bytes)
</pre>
 
<dt><b><a name="dbadmin">Dbadmin</a></b>
 
    <dd> James Harrell
    <a href="mailto:gt4960a@prism.gatech.edu">
    &lt;gt4960a@prism.gatech.edu&gt;</a> has developed a CGI
    application that allows database administration using a web based
    form interface. A demonstration can be seen at: <p>
 
<pre>
    <a href="http://bauhaus.skiles.gatech.edu/~jharrell">http://bauhaus.skiles.gatech.edu/~jharrell</a>
</pre>
 
    Source code is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/dbadmin/dbadmin_v1.0.1.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/dbadmin/dbadmin_v1.0.1.tar.gz</a>
</pre>
 
 
<dt><b><a name="dbase">DBASE</a></b>
 
    <dd> Maarten Boekhold
    <a href="mailto:M.Boekhold@et.tudelft.nl">
    &lt;M.Boekhold@et.tudelft.nl&gt;</a> has written a dbase to mSQL
    conversion utility. It handles the dbf data types <tt>string</tt>,
    <tt>num</tt>, <tt>real</tt> and <tt>date</tt>. It does not yet
    handle <tt>memo-fields</tt> because mSQL 1.<i>x</i> lacks support
    for variable length <tt>char</tt> fields. Maarten is aware that
    this may be insufficient for some users' needs but he is releasing
    the utility as 'do-whatever-you-want' software. It is available via
    anonymous ftp from:
 
<pre>
    <a href="ftp://ftp.nerosworld.com/pub/SQL/dbf2sql/dbf2sql-2.2.tar.gz">ftp://ftp.nerosworld.com/pub/SQL/dbf2sql/dbf2sql-2.2.tar.gz</a>
    (18207 bytes)
</pre>
 
<dt><b><a name="dbi">DBI/DBD</a></b>
 
    <dd> Tim Bunce
    <a href="mailto:Tim.Bunce@ig.co.uk">&lt;Tim.Bunce@ig.co.uk&gt;</a>
    is working on a generic database driver (DBI) which allows perl to
    interface to any database in a standardised way. Details of Tim's
    work are available from: <p>
 
<pre>
    <a href="http://www.hermetica.com">http://www.hermetica.com</a>
</pre>
 
    <a name="dbd-msql"><dd> Alligator Decartes</a>
    <a href="mailto:descarte@hermetica.com">
    &lt;descarte@hermetica.com&gt;</a> has added an mSQL driver to
    Tim's work and his contributions can be obtained via anonymous ftp
    from: <p>
 
<pre>
    <a href="ftp://ftp.mcqueen.com/pub/dbperl">ftp://ftp.mcqueen.com/pub/dbperl</a>
</pre>
 
    It is also available from any CPAN (Comprehensive Perl Archive
    Network) site in the &quot;modules&quot; directory. For more
    information about CPAN see: <p>
 
<pre>
    <a href="ftp://ftp.funet.fi/pub/languages/perl/CPAN">ftp://ftp.funet.fi/pub/languages/perl/CPAN</a>
</pre>
 
    The latest blurb describing Alligator's work can be obtained from: <p>
 
<pre>
    <a href="http://www.hermetica.com/technologia/DBI">http://www.hermetica.com/technologia/DBI</a>
</pre>
 
<dt><b><a name="dbunk">DBunk</a></b>
 
    <dd> Dave Shevett <a href="mailto:shevett@homeport.org">
    &lt;shevett@homeport.org&gt;</a> has written DBunk - a Java based
    graphical front end to mSQL. Source code is available from: <p>
 
<pre>
    <a href="http://www.homeport.org/~shevett/dbunk.tar.gz">http://www.homeport.org/~shevett/dbunk.tar.gz</a>
</pre>
 
<dt><b><a name="dbview">Dbview</a></b>
 
    <dd> Gian Paolo Ciceri
    <a href="mailto:gp.ciceri@it.net">&lt;gp.ciceri@it.net&gt;</a>
    has written a utility similar to
    <a href="#support-progs">relshow</a>. <p>
 
    It shows the structure of an mSQL database as well as indicating
    the number of records in the tables. <p>
 
    It is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/dbview/dbview.c.lsm">ftp://bond.edu.au/pub/Minerva/msql/Contrib/dbview/dbview.c.lsm</a>
    (563 bytes)
 
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/dbview/dbview.c">ftp://bond.edu.au/pub/Minerva/msql/Contrib/dbview/dbview.c</a>
    (16933 bytes)
</pre>
 
<dt><b><a name="digger">Digger</a></b>
 
    <dd> The folks at
    <a href="http://www.bunyip.com">Bunyip Information Services</a>
    (the current maintainers of the <a href="#mailing-list">mSQL
    mailing list</a>) have used mSQL as the database component of their
    Digger system. <p>
 
    Digger is a Distributed Directory Service for the Internet based on
    Whois++ technology. For more information about digger send mail to
    <a href="mailto:digger-info@bunyip.com">
    &lt;digger-info@bunyip.com&gt;</a> or have a look at Bunyip's web
    pages:
 
<pre>
    <a href="http://www.bunyip.com/products/digger">http://www.bunyip.com/products/digger</a>
</pre>
 
<dt><b><a name="emacs">Emacs</a></b>
 
    <dd> Igor Romanenko
    <a href="mailto:igor@frog.kiev.ua">&lt;igor@frog.kiev.ua&gt;</a>
    has contributed some lisp code to provide emacs with hooks into
    mSQL. It allows the msql monitor to run in an emacs window, so you
    can use emacs for editing and command recall. It is available via
    anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLEmacs/sql-mode.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLEmacs/sql-mode.tar.gz</a>
    (6883 bytes)
</pre>
 
<dt><b><a name="msql-import">Flat File Importer</a></b>
 
    <dd> Pascal Forget
    <a href="mailto:pascal@wsc.com">&lt;pascal@wsc.com&gt;</a> has
    contributed a program that will import flat file databases directly
    into mSQL databases. It supports both the version 1.<i>x</i> and
    2.<i>x</i> releases of mSQL. It is available via anonymous ftp
    from: <p>
 
<pre>
    <a href="ftp://ftp.wsc.com/pub/freeware/msql/msql-import-0.1.4.tar.gz">ftp://ftp.wsc.com/pub/freeware/msql/msql-import-0.1.4.tar.gz</a>
    (14281 bytes)
</pre>
 
    Notwithstanding any bug fixes, this will be the last
    &quot;stand alone&quot; release of <tt>msql-import</tt> that can be
    used with mSQL version 1.<i>x</i>. Future versions will be bundled
    with mSQL version 2.<i>x</i>. <p>
 
<dt><b><a name="html-example">HTML Interface to mSQL</a></b>
 
    <dd> Sol Katz
    <a href="mailto:skatz@blm.gov">&lt;skatz@blm.gov&gt;</a> has
    written some C code that demonstrates an HTML interface to mSQL.
    Unlike other applications, this program does
    not need to be modified when new mSQL tables or
    databases are created. All required information is
    contained in the html. <p>
 
    It is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLCGI/msqlc2.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLCGI/msqlc2.zip</a>
    (29002 bytes)
</pre>
 
    and also from: <p>
 
<pre>
    <a href="ftp://ftp.blm.gov/pub/gis/msqlc2.zip">ftp://ftp.blm.gov/pub/gis/msqlc2.zip</a>
    (29002 bytes)
</pre>
 
    It can be seen in operation at: <p>
 
<pre>
    <a href="http://www.blm.gov/gis/msql/vertical/test2.html">http://www.blm.gov/gis/msql/vertical/test2.html</a>
</pre>
 
<dt><b><a name="ici">ICI</a></b>
 
    <dd> Yiorgos Adamopoulos
    <a href="mailto:Y.Adamopoulos@noc.ntua.gr">
    &lt;Y.Adamopoulos@noc.ntua.gr&gt;</a> has written an mSQL extension
    for the ICI programming language. It is available via anonymous ftp
    from: <p>
 
<pre>
    <a href="ftp://ftp.ntua.gr/pub/lang/ici/iciMsql.tar.gz">ftp://ftp.ntua.gr/pub/lang/ici/iciMsql.tar.gz</a>
    (249077 bytes)
</pre>
 
    Yiorgos has set up a mailing list for <tt>iciMsql</tt>. Questions
    can be sent to: <p>
 
<pre>
    <a href="mailto:iciMsql@noc.ntua.gr">iciMsql@noc.ntua.gr</a>
</pre>
 
<dt><b><a name="jate">Jate</a></b>
 
    <dd> Josef <a href="mailto:ht@aiace.lnf.infn.it">
    &lt;ht@aiace.lnf.infn.it&gt;</a> has developed Jate - a CGI
    program that builds HTML interfaces to mSQL databases. More
    details can be found at: <p>
 
<pre>
    <a href="http://aiace.lnf.infn.it/~ht/JATE.html">http://aiace.lnf.infn.it/~ht/JATE.html</a>
</pre>
 
<dt>(-) <b><a name="java">Java</a></b>
 
    <dd> Darryl Collins
    <a href="mailto:darryl@minmet.uq.oz.au">
    &lt;darryl@minmet.uq.oz.au&gt;</a> has developed a version of the
    mSQL API for the Java programming language. <p>
 
    MsqlJava for mSQL version 1.<i>x</i> is available via anonymous
    ftp from: <p>
 
<pre>
    <a href="ftp://dada.minmet.uq.edu.au/pub">ftp://dada.minmet.uq.edu.au/pub</a>
</pre>
 
    MsqlJava for mSQL version 2.<i>x</i>  is available via anonymous
    ftp from: <p>
 
<pre>
    <a href="ftp://dada.minmet.uq.edu.au/pub/MsqlJava-1.2.0.tar.gz">ftp://dada.minmet.uq.edu.au/pub/MsqlJava-1.2.0.tar.gz</a>
    (23483 bytes)
</pre>
 
    For more details see: <p>
 
<pre>
    <a href="http://www.minmet.uq.oz.au/msqljava">http://www.minmet.uq.oz.au/msqljava</a>
</pre>
 
    <dd> George Reese <a href="mailto:borg@imaginary.com">
    &lt;borg@imaginary.com&gt;</a> has developed a Java Database
    Connection (JDBC) class library using MsqlJava. It is available via
    anonymous ftp from: <p>
 
 
<pre>
    <a href="ftp://ftp.imaginary.com/pub/Java/database/mSQL-JDBC.tar.gz">ftp://ftp.imaginary.com/pub/Java/database/mSQL-JDBC.tar.gz</a>
    (33063 bytes)
</pre>
 
    For more details see: <p>
 
<pre>
    <a href="http://www.imaginary.com/Java">http://www.imaginary.com/Java</a>
</pre>
 
    <dd> Douglas Beeson <a href="mailto:beesond@ecworld.org">
    &lt;beesond@ecworld.org&gt;</a> reports:
 
<pre>
    Sun has just released its Java Servlets API that includes, among other
    things, a Java runtime that works from within Netscape servers, similar
    to a plug-in. The same servlet runner is available for ISAPI and the
    Apache server. With a servlet written in Java and using the Java-mSQL
    package cited in the FAQ, you can open a connection to an mSQL database
    on the local or remote server, query the DB and return the result as
    HTML using an http response object pre-defined in the servlet API.
    Best of all, the servlets API is free.
 
    <a href="http://jserv.javasoft.com/products/java-server/sdk/index.shtml">http://jserv.javasoft.com/products/java-server/sdk/index.shtml</a>
</pre>
 
<dt><b><a name="jio">Jio</a></b>
 
    <dd> Simon Kenyon <a href="mailto:S.Kenyon@koala.ie">
    &lt;S.Kenyon@koala.ie&gt;</a> has developed Jio which is a dynamic
    web content generation system written as a Java servlet. It uses
    JDBC to connect to mSQL which contains the data. All generated
    pages are created using templates. Source code is included and it
    is free for commercial use. See the COPYRIGHT file included with
    the Jio distribution for details. <p>
 
    It is available via anonymous ftp from:  <p>
 
<pre>
    <a href="http://www.koala.ie/jio/dist/jio-1.3.tar.gz">http://www.koala.ie/jio/dist/jio-1.3.tar.gz</a>
    (35104 bytes)
</pre>
 
<dt><b><a name="siod">LISP/Scheme</a></b>
 
    <dd> George J. Carrette <a href="mailto:gjc@world.std.com">
    &lt;gjc@world.std.com&gt;</a> has developed SIOD which is
    &quot;a small-footprint implementation of the Scheme programming
    language that is provided with some database, unix
    programming and cgi scripting extensions&quot;. SIOD provides
    interfaces to Oracle, Sybase, mSQL and Digital RDB databases. <p>
 
    Details of the current release of SIOD can be found at: <p>
 
<pre>
    <a href="http://people.delphi.com/gjc/siod.html">http://people.delphi.com/gjc/siod.html</a>
</pre>
 
    The latest copy of SIOD can usually be found at: <p>
 
<pre>
    <a href="ftp://ftp.std.com/pub/gjc/siod.tgz">ftp://ftp.std.com/pub/gjc/siod.tgz</a>
    (156846 bytes)
</pre>
 
    or
 
<pre>
    <a href="http://people.delphi.com/gjc/siod.tgz">http://people.delphi.com/gjc/siod.tgz</a>
    (156730 bytes)
</pre>
 
    George has used SIOD as part of a system to: <p>
 
<pre>
    abstract html tricks from programming tricks by considering html
    files, or portions of html files, to be templates or chunks of html
    upon which simple substitutions are made
</pre>
 
    More details of &quot;Chunk HTML&quot; can be found at: <p>
 
<pre>
    <a href="http://people.delphi.com/gjc/chtml.html">http://people.delphi.com/gjc/chtml.html</a>
</pre>
 
<dt><b><a name="literepository">Lite Repository</a></b>
 
    <dd> Jeanne Gold <a href="mailto:goldj@Ix.netcom.com">
    &lt;goldj@Ix.netcom.com&gt;</a> has established a site for Lite
    contributions. For more details see: <p>
 
<pre>
    <a href="http://www.gold-cousins.org/lite_site">http://www.gold-cousins.org/lite_site</a>
</pre>
 
<dt><b><a name="metahtml">Meta-HTML</a></b>
 
    <dd> Universal Access Inc has released a version of their Meta-HTML
    scripting language which has extensions for connecting to mSQL
    databases. <p>
 
    Henry Minsky <a href="mailto:hqm@ua.com">
    &lt;hqm@ua.com&gt;</a> writes: <p>
<pre>
    &lt;Meta-HTML&gt; is a programming language specifically designed
    for working within the World Wide Web environment. Although it is a
    genuine programming language, suitable for large-scale symbolic
    manipulation, it provides the most commonly wanted Web
    functionality as built-in primitives, so you don't have to write
    them.
</pre>
 
    More details (including source code) can be found at: <p>
 
<pre>
    <a href="http://www.metahtml.com">http://www.metahtml.com</a>
</pre>
 
    There is also an anonymous ftp site for those of you without web
    access: <p>
 
<pre>
    <a href="ftp://ftp.metahtml.com/pub">ftp://ftp.metahtml.com/pub</a>
</pre>
 
<dt><b><a name="ms-access7">MS Access to mSQL</a></b>
 
    <dd> Brian Andrews <a href="mailto:brian@informate.co.nz">
    &lt;brian@informate.co.nz&gt;</a> has written a function for use
    with MS Access version 7. This function will take an Access 7
    database and generate the appropriate mSQL <tt>INSERT</tt>
    statements for importing the Access 7 data into an mSQL database.
    <p>
 
    Brian's original function is available from the mSQL mailing list
    <a href="#list-archives">archives</a> for the month of January
    1997. <p>
 
    Jon Hilton <a href="mailto:jon@hisa.org.au">
    &lt;jon@hisa.org.au&gt;</a> has taken Brian's work and corrected a
    few bugs and included support for mSQL version 2.<i>x</i>. <p>
 
    Jon's function is available from the mSQL mailing list
    <a href="#list-archives">archives</a> for the month of February
    1997. It is also included here in full: <p>
 
<pre>
<font size="-1">
Function export_mSQL()
 
    ' Exports the database contents into a file in mSQL format
    ' IS NOT SELECTIVE! (exports ALL tables)
 
    Dim dbase As DATABASE, tdef As Recordset, i As Integer, fd As Integer
 
    Set dbase = CurrentDb()
 
    ' Open the export file
    Open &quot;E:\cota\infocom1.txt&quot; For Output As #1
 
    Print #1, &quot;# Converted from MS Access to mSQL &quot;
    Print #1, &quot;# by Brian Andrews, (c) InforMate Technologies, 1997&quot;
    Print #1, &quot;&quot;
 
    ' Go through the table definitions
 
 
    For i = 0 To dbase.TableDefs.Count - 1
        Print #1, &quot;# TableDefs.Count - 1 =&quot; &amp; dbase.TableDefs.Count - 1
        Print #1, &quot;# This is table &quot; &amp; i
        ' Let's take only the visible tables
 
        If ((dbase.TableDefs(i).Attributes And DB_SYSTEMOBJECT) Or
(dbase.TableDefs(i).Attributes And DB_HIDDENOBJECT)) Then
 
        Else
 
            ' We DROP the table in case it already exists
            ' and then create it again
 
                    tname = &quot;&quot; &amp; dbase.TableDefs(i).Name
                    Print #1, &quot;# Access table &quot; &amp; tname
 
            'remove spaces from tablename
 
                        For j = 1 To Len(tname)
 
                        If j &lt; Len(tname) Then
 
                        If Mid$(tname, j, 1) = &quot; &quot; Then
 
 
                        s = Left$(tname, j - 1)
                        's = s &amp; &quot;&quot; &amp; Right$(t.Fields(j), Len(t.Fields(j)) - i + 1)
                        s = s &amp; &quot;&quot; &amp; Right$(tname, Len(tname) - j)
                        j = j + 1
                        found = True
                        tname = s
                        'Exit For
 
                        End If
                        End If
 
                        Next j
 
            'restrict tablename to 19 chars
 
            tname = Left$(tname, 19)
 
            Print #1, &quot;&quot;
            Print #1, &quot;&quot;
            Print #1, &quot;DROP TABLE &quot; &amp; tname &amp; &quot; \p\g&quot;
            Print #1,
            Print #1, &quot;CREATE TABLE &quot; &amp; tname &amp; &quot;(&quot;
 
 
            ' Step through all the fields in the table
 
            For fd = 0 To dbase.TableDefs(i).Fields.Count - 1
 
            'All fields are char at the moment - can be changed
 
                Dim tyyppi As String, pituus As Integer, comma As String
                Select Case dbase.TableDefs(i).Fields(fd).Type
                    Case DB_BOOLEAN
                        tyyppi = &quot;char (8)&quot;
                    Case DB_INTEGER, DB_BYTE, DB_LONG
                        tyyppi = &quot;int&quot;
                    Case DB_DOUBLE, DB_SINGLE, DB_CURRENCY
                        tyyppi = &quot;real&quot;
                    Case DB_TEXT
                        pituus = dbase.TableDefs(i).Fields(fd).Size
                        tyyppi = &quot;char (&quot; &amp; pituus &amp; &quot;)&quot;
               ' Need to leave enough room in date fields for date &amp; time
                    Case DB_DATE
                         tyyppi = &quot;char (17)&quot;
               ' For some reason, DB_GUID doesn't seem to be noticed - 15 is the value
                    Case DB_MEMO, DB_GUID, DB_LONGBINARY, 15
 
               'Special case fields - specific to one of my projects
 
                        If dbase.TableDefs(i).Fields(fd).Name = &quot;Fund Summary &quot; Then
                        tyyppi = &quot;char (1500)&quot;
                        ElseIf dbase.TableDefs(i).Fields(fd).Name = &quot;Fund Analysis &quot; Then
                        tyyppi = &quot;char (3000)&quot;
                        Else
                        tyyppi = &quot;text (100)&quot;
                        End If
 
                End Select
 
                ' Don't print the separating comma after the last field
 
                If fd &lt; dbase.TableDefs(i).Fields.Count - 1 Then
                    comma = &quot;,&quot;
                Else
                    comma = &quot;&quot;
                End If
 
                ' Print the field definition
 
                'remove spaces from fieldname
 
                       stuff = &quot;&quot; &amp; dbase.TableDefs(i).Fields(fd).Name
 
                        For j = 1 To Len(stuff)
 
                        If j &lt; Len(stuff) Then
 
                        If Mid$(stuff, j, 1) = &quot; &quot; Then
 
 
                        s = Left$(stuff, j - 1)
                        's = s &amp; &quot;&quot; &amp; Right$(t.Fields(j), Len(t.Fields(j)) - i + 1)
                        s = s &amp; &quot;&quot; &amp; Right$(stuff, Len(stuff) - j)
                        j = j + 1
                        found = True
                        stuff = s
                        'Exit For
 
                        End If
                        End If
 
                        Next j
 
               stuff = Left$(stuff, 19)
 
               'mSQL 1 primary key declaration - always on first field
               'mSQL 2 - now we need to use CREATE INDEX. Don't bother for now.
 
               'If fd = 0 Then
                'Print #1, &quot;     &quot; &amp; stuff &amp; &quot; &quot; &amp; tyyppi &amp; &quot; primary key&quot; &amp; comma
                'Else
                Print #1, &quot;     &quot; &amp; stuff &amp; &quot; &quot; &amp; tyyppi &amp; comma
               'End If
 
            Next fd
 
            Print #1, &quot;)\p\g&quot;
            Print #1, &quot;&quot;
 
 
            Dim recset As Recordset
            Set recset = dbase.OpenRecordset(dbase.TableDefs(i).Name)
 
            ' Step through the rows in the table
            reccount = recset.RecordCount
            If reccount &lt;&gt; 0 Then
            recset.MoveFirst
            Do Until recset.EOF
                Dim row As String, it As String
 
                row = &quot;INSERT INTO &quot; &amp; tname &amp; &quot; VALUES (&quot;
 
                ' Go through the fields in the row
 
                For fd = 0 To recset.Fields.Count - 1
                    Dim is_string As String
 
                    is_string = &quot;&quot;
                    stuff = &quot;&quot; &amp; recset.Fields(fd).Value
                    Select Case recset.Fields(fd).Type
              ' For some reason, DB_GUID doesn't seem to be noticed - 15 is the value
                    Case DB_TEXT, DB_MEMO, DB_GUID, DB_DATE, DB_LONGBINARY, DB_BOOLEAN, 15
                        is_string = &quot;'&quot;
                    Case Else
                         If stuff = &quot;&quot; Then
                            stuff = &quot;0&quot;
                         End If
                    End Select
 
 
 
 
                    '**** escape single quotes
                    x = InStr(stuff, &quot;'&quot;)
                    While x &lt;&gt; 0
                            s = Left$(stuff, x - 1)
                            s = s &amp; &quot;\&quot; &amp; Right$(stuff, Len(stuff) - x + 1)
                            stuff = s
                            x = InStr(x + 2, stuff, &quot;'&quot;)
                    Wend
 
                    row = row &amp; is_string &amp; stuff &amp; is_string
                    If fd &lt; recset.Fields.Count - 1 Then
                        row = row &amp; &quot;,&quot;
                    End If
                Next fd
 
                ' Add trailers and print
 
                row = row &amp; &quot;)\p\g&quot;
                Print #1, row
 
                ' Move to the next row
 
                recset.MoveNext
            Loop
 
            recset.Close
            Set recset = Nothing
 
        End If
    End If
    Next i
 
 
 
    Close #1
 
    dbase.Close
    Set dbase = Nothing
 
End Function
</font>
</pre>
 
<dt><b><a name="ms-access-dbase3">MS Access &amp; dBase III</a></b>
 
    <dd> Brian Bartholemew
    <a href="mailto:bb@wizard.pn.com">&lt;bb@wizard.pn.com&gt;</a>
    has written some scripts that translate dBase III tables saved from
    MS Access into tab-separated formats. <p>
 
    Brian writes: <p>
 
<pre>
    Here are three scripts, the first two are clean ones that break out
    tables saved from access in dBaseIII format into a tab-separated
    format that a set of database-operator scripts called rdb can use.
    The first breaks out .dbf files and the second breaks out .dbt
    files which contain the text of variable-length-text memo fields;
    the .dbf and .dbt can be joined by the block offset number given in
    the text field. These scripts have only been tested on the field
    types my database happens to have. Since the first one breaks out
    the table definitions it's the obvious candidate to generate table
    definitions for msql. The third script is a hack to get data into
    msql so I can play with msql, but it's a starting point. The first
    two scripts were written to be free of data-dependent bugs, the
    third is somehow confused about null fields in Pascal's msql-import
    program which I've hacked around to enter the complaining fields as
    the text &quot;NULL&quot;; debugging is welcomed. The third script
    needs rdb, available from <a href="ftp://rand.org/pub/RDB-hobbs">rand.org:/pub/RDB-hobbs</a>. I do MS Access
    -&gt; rdb -&gt; msql because I use rdb as a prototyping tool, the
    format is trivial to generate and modify, I have a forms-based
    emacs front-end to rdb, and the tables compress nicely in rcs.
    However, someone may wish to modify these scripts, (a) so that they
    generate msql dump files instead of rdb files, thereby bypassing
    the msql-import bug and the rdb dependency and probably some data
    dependencies with maximum portability, or (b) so that they talk to
    the database directly. If so please post diffs. Trigger the new
    behaviour by a command-line option so the non-msql-perl behaviour
    continues to run under vanilla perl. These programs are gpl'ed.
</pre>
 
    They are available from the mSQL mailing list
    <a href="#list-archives">archives</a> for the month of February
    1996. <p>
 
<dt><b><a name="msqlcgi">mSQL CGI</a></b>
 
    <dd> Alex Tang <a href="mailto:altitude@petrified.cic.net">
    &lt;altitude@petrified.cic.net&gt;</a> has written an mSQL front
    end as a CGI program. For more details see: <p>
 
<pre>
    <a href="http://petrified.cic.net/MsqlCGI">http://petrified.cic.net/MsqlCGI</a>
</pre>
 
<dt><b><a name="msqlexpire">mSQLexpire</a></b>
 
    <dd> Scott Burkett <a href="mailto:scottb@dcicorp.com">
    &lt;scottb@dcicorp.com&gt;</a> has developed an mSQL utility that
    provides a mechanism for automatic expiration of records based on
    age. Further details can be seen at:  <p>
 
<pre>
    <a href="http://www.dcicorp.com/~scottb/projects/msqlexpire">http://www.dcicorp.com/~scottb/projects/msqlexpire</a>
</pre>
 
<dt><b><a name="msqlexport">mSQL-Export</a></b>
 
    <dd> Kerry Garrison <a href="mailto:garrison@delta.net">
    &lt;garrison@delta.net&gt;</a> has developed a Perl script that
    will export an mSQL table to a delimited text file. Further
    details can be seen at: <p>
 
<pre>
    <a href="http://design.delta.net/msqltools">http://design.delta.net/msqltools</a>
</pre>
 
<dt><b><a name="msqlkeeper">mSQL Keeper</a></b>
 
    <dd> Chris Lambrou <a href="mailto:chrislambrou@ineternetmci.com">
    &lt;chrislambrou@ineternetmci.com&gt;</a> has written an mSQL
    maintenance utility called mSQL Keeper. <p>
 
    &quot;mSQL Keeper is a collection of perl scripts that enables easy
    management of an mSQL database. It requires no knowledge of SQL.
    All database maintenance functions are executed through a GUI front
    end, and SQL statements are built and submitted to the mSQL engine
    from mSQL Keeper.&quot; <p>
 
    mSQL Keeper currently supports mSQL 1.<i>x</i> databases. Chris has
    plans for an mSQL 2.<i>x</i> version once mSQL 2.<i>x</i>
    stabilises. <p>
 
    More details can be found at: <p>
 
<pre>
    <a href="http://www.cglis.com/msql_keeper">http://www.cglis.com/msql_keeper</a>
</pre>
 
<dt><b><a name="msqlsummary">mSQL Summary</a></b>
 
    <dd> David Perry <a href="mailto:deperry@nerosworld.com">
    &lt;deperry@nerosworld.com&gt;</a> had someone write a program for
    him that takes an SQL select statement (such as select distinct
    fieldname, fieldname2, fieldnameN from tablename order by
    fieldname) and propagates a second table with the results. It is
    available via anonymous ftp from:  <p>
 
<pre>
    <a href="ftp://ftp.nerosworld.com/pub/msql/Contrib/mSQLsummary/mSQL_summary.tar">ftp://ftp.nerosworld.com/pub/msql/Contrib/mSQLsummary/mSQL_summary.tar</a>
</pre>
 
<dt><b><a name="msqltools">mSQL Tools</a></b>
 
    <dd> Kerry Garrison <a href="mailto:garrison@delta.net">
    &lt;garrison@delta.net&gt;</a> has setup a site that catalogues a
    number of mSQL tools. It can be found at: <p>
 
<pre>
    <a href="http://design.delta.net/msqltools">http://design.delta.net/msqltools</a>
</pre>
 
<dt><b><a name="msqlui">mSQL User Interface for Windows</a></b>
 
    <dd> Chris Mai <a href="mailto:chrissde@aol.com">
    &lt;chrissde@aol.com&gt;</a> has developed a user interface for
    Microsoft Windows that will connect to mSQL databases. It is
    available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLui/msql-ui.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLui/msql-ui.zip</a>
</pre>
 
<dt>(=) <b><a name="msutil">msutil - an alternative C API for mSQL</a></b>
 
    <dd> Rich Bryant
    <a href="mailto:rbb@iseek.com">
    &lt;rbb@iseek.com&gt;</a> has developed an alternate C API for use
    with mSQL. Details on msutil can be found at: <p>
 
<pre>
    <a href="http://iseek.com/biztek/docs/msutil">http://iseek.com/biztek/docs/msutil</a>
</pre>
 
<dt><b><a name="ms-windows">MS Windows</a></b>
 
    <dd> Dean Fuqua
    <a href="mailto:fuqua@niehs.nih.gov">
    &lt;fuqua@niehs.nih.gov&gt;</a> has contributed an mSQL API for MS
    Windows platforms. To use this software you need some Winsock
    compliant stack. (i.e. Trumpet Winsock, MS TCP/IP-32, Chameleon,
    etc.) It is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLwinapi/winapi.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLwinapi/winapi.zip</a>
    (87211 bytes)
</pre>
 
    Not included in <tt>winapi.zip</tt> is an <tt>msql.ini</tt> file.
    Its contents should resemble <p>
 
<pre>
    [Server]
    IP=your.server.host.name
    Port=1112
    Username=YourUsername
</pre>
 
    There also appears to be a later version of Dean's work which
    includes compiled executables in <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/Win-mSQL/winmsql7.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/Win-mSQL/winmsql7.zip</a>
    (306827 bytes)
</pre>
 
    Peter Tillemans <a href="mailto:pti@net4all.be">
    &lt;pti@net4all.be&gt;</a> has taken Dirk Ohme's
    <a href="mailto:dohme@transtec.de"> &lt;dohme@transtec.de&gt;</a>
    OS/2 port of mSQL and used it to produce a Windows 95/NT version of
    mSQL. This port can be obtained via anonymous ftp from:  <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/Win-mSQL/msql116b-w32.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/Win-mSQL/msql116b-w32.zip</a>
    (755355 bytes)
</pre>
 
    Peter Tillemans <a href="mailto:pti@net4all.be">
    &lt;pti@net4all.be&gt;</a> and Fernando Lozano
    <a href="mailto:bl@blnet.com"> &lt;bl@blnet.com&gt;</a> have
    created the &quot;mSQL PC Home-Page&quot;. Their work can be seen
    at: <p>
 
<pre>
    <a href="http://blnet.com/msqlpc">http://blnet.com/msqlpc</a>
</pre>
 
<dt><b><a name="nextstep">NeXTSTEP EOF</a></b>
 
    <dd> Mark Onyschuk
    <a href="mailto:ask-oa@oa.guild.org">
    &lt;ask-oa@oa.guild.org&gt;</a> has developed an NeXTSTEP EOF
    adaptor for mSQL. <p>
 
    An Enterprise Object Framework (EOF) is an object framework that
    allows object oriented access to relational databases, where each
    row is considered an object. Besides a few limitations, it
    basically makes a relational database look like an OO database to
    the developer. By means of an adaptor, EOF can be used with
    virtually any database. The adaptor is responsible to transform the
    generic OO messages in database specific queries by subclassing a
    generic adaptor and modifying its behaviour. <p>
 
<dt><b><a name="oo-html">Object Oriented HTML API</a></b>
 
    <dd> Sol Katz
    <a href="mailto:skatz@blm.gov">&lt;skatz@blm.gov&gt;</a> has
    developed an object oriented HTML API for mSQL. <p>
 
    It is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.blm.gov/pub/gis/msql_api.tar.gz">ftp://ftp.blm.gov/pub/gis/msql_api.tar.gz</a>
    (10317 bytes)
</pre>
 
    An example can be found at
 
<pre>
    <a href="http://www.blm.gov/gis/msql/dbs6.html">http://www.blm.gov/gis/msql/dbs6.html</a>
</pre>
 
<dt><b><a name="ocx">OCX</a></b>
 
    <dd> Shayne Hughes
    <a href="mailto:shayne@vgl.ucdavis.edu">
    &lt;shayne@vgl.ucdavis.edu&gt;</a> has created a 32-bit OCX control
    that exposes all of the mSQL API (based on 1.0.12) with only slight
    modifications to make it work with pointer impaired languages like
    Visual Basic. It is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://vgl.ucdavis.edu/pub/mSQL/ocxmsql-0.90.zip">ftp://vgl.ucdavis.edu/pub/mSQL/ocxmsql-0.90.zip</a>
    (1685189 bytes)
</pre>
 
    Chin-Jin Phua
    <a href="mailto:jojo@ttsh.gov.sg">
    &lt;jojo@ttsh.gov.sg&gt;</a> has also developed a mSQL 32-bit OCX
    for Windows95 and WinNT. He has only tested it for Visual Basic.
 
<pre>
    <a href="ftp://Bond.edu.au/pub/Minerva/msql/Contrib/mSQLwinapi/msqlocx.zip">ftp://Bond.edu.au/pub/Minerva/msql/Contrib/mSQLwinapi/msqlocx.zip</a>
    (17409 bytes)
</pre>
 
 
<dt><b><a name="odbc">ODBC</a></b>
 
<dd>
<ul>
    <li>Dean Fuqua <a href="mailto:fuqua@niehs.nih.gov">
    &lt;fuqua@niehs.nih.gov&gt;</a> has developed a collection of ODBC
    files which are available via anonymous ftp from:  <p>
 
<pre>
    <a href="ftp://ftp.comed.com/pub/msql/odbc">ftp://ftp.comed.com/pub/msql/odbc</a>
</pre>
 
    For more information on Dean's work see:
 
<pre>
    <a href="http://alfred.niehs.nih.gov">http://alfred.niehs.nih.gov</a>
</pre>
 
    <li> Kevin Gill <a href="mailto:kgill@kindle.ie">
    &lt;kgill@kindle.ie&gt;</a> has extended Dean's work and developed
    ODBC software for connecting Visual Basic 3.0 to an mSQL server.
    His work can be obtained via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC</a>
</pre>
 
    The following files give some out of date general information about
    the gorta software:  <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gorta.faq">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gorta.faq</a>
    (27450 bytes)
 
    <a href="http://www.cyber.com.au/misc/gorta.htm">http://www.cyber.com.au/misc/gorta.htm</a>
</pre>
 
    The gorta software operates as either a 32 bit or a 16 bit ODBC driver.
    Different libraries/programs are required for each. <p>
 
    For the 16 bit version (Windows 3.1 and 3.11) install: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmsqlrts.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmsqlrts.zip</a>
    (11693 bytes)	- Mini-SQL API DLL (Must install First)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gorta-rts.2.10.0.6.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gorta-rts.2.10.0.6.zip</a>
    (17014 bytes)	- Gorta Driver
</pre>
 
    For the 32 bit version (Windows 95) install: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmsqlr32-rts.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmsqlr32-rts.zip</a>
    (43193 bytes)	- Mini-SQL API 32 bit DLL (Must Install First)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gorta32-rts.2.10.0.6.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gorta32-rts.2.10.0.6.zip</a>
    (22260 bytes)	- Gorta 32 bit driver
</pre>
 
    The following test tools are available. (These are 16 bit but should
    work with the 32 bit drivers). <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmslqrs.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmslqrs.zip</a>
    (20257 bytes)	- A DOS/Windows version of the relshow program
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gortars.1.0.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gortars.1.0.zip</a>
    (23756 bytes)	- An ODBC version of the relshow program
</pre>
 
    To install the ODBC driver for the first time requires the use of an
    install disk.  This installs version 2.10.0.3 of the gorta drivers.
    Install from this and then overwrite the gorta dll (in the
    windows/system directory) with the one in the gorta 2.10.0.6 zip
    file.  The install disk installs both the 16 and 32 bit versions of
    ODBC.  <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/installd.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/installd.zip</a>
    (225019 bytes)
</pre>
 
    The install disk tends to replace dlls with older versions.  This is
    a problem if you have ODBC already running.  Check the DLLs on the
    disk, and make a copy of your dlls before starting.  <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gortars.1.0.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gortars.1.0.zip</a>
    (23756 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gortarts.2.10.0.3.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gortarts.2.10.0.3.zip</a>
    (17480 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gortasrc.2.10.0.3.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/gortasrc.2.10.0.3.zip</a>
    (51770 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/installd.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/installd.zip</a>
    (225019 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/msql-1.12-dump.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/msql-1.12-dump.tar.gz</a>
    (158282 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmsqlrs.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmsqlrs.zip</a>
    (20257 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmsqlrts.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmsqlrts.zip</a>
    (11693 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmsqlsrc.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLODBC/wmsqlsrc.zip</a>
    (15089 bytes)
</pre>
 
    <li> <a name="iODBC">An</a> ODBC API for Linux is available. For more
    details see: <p>
 
<pre>
    <a href="http://users.ids.net/~bjepson/freeODBC">http://users.ids.net/~bjepson/freeODBC</a>
</pre>
 
    Dirk Ohme <a href="mailto:dohme@transtec.de">
    &lt;dohme@transtec.de&gt;</a> has ported this API to OS/2, SunOS
    4.<i>x</i> and Solaris 2.<i>x</i>. This port is available via
    anonymous ftp from:
 
<pre>
    <a href="ftp://ftp.nerosworld.com/pub/SQL/iODBC/iODBC-1.00.5.zip">ftp://ftp.nerosworld.com/pub/SQL/iODBC/iODBC-1.00.5.zip</a>
    (1023149 bytes)
</pre>
 
    <li> Ian Nicholls <a href="mailto:iann@ram.net.au">
    &lt;iann@ram.net.au&gt;</a> of Ram Network Services has developed
    a Delphi class that will interface with mSQL version 1.<i>x</i>.
    Ian also reports that they have a C++ version which runs under
    Linux. He stresses that the code is still only in its beta test
    phase but is &quot;getting stronger every day&quot;. <p>
 
    For more details please send some mail to Ian at
    <a href="mailto:iann@ram.net.au"> &lt;iann@ram.net.au&gt;</a>.
 
</ul>
<p>
 
<dt><b><a name="onyx">Onyx</a></b>
 
    <dd> Michael Koehne
    <a href="mailto:kraehe@bakunin.north.de">
    &lt;kraehe@bakunin.north.de&gt;</a> has developed a rapid
    prototyping tool for database applications called Onyx. <p>
 
    Onyx consists of a transaction manager, a shell like 4GL and a
    Simple Database Transaction Protocol engine. <p>
 
    &quot;Onyx is designed by the Model-View-Controller paradigm, so
    tables are the model, masks are the views and transactions are the
    controllers which can be bound to an input field, a menu, function
    keys or the change of the current record in a cursor.&quot; <p>
 
    It is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.uni-bremen.de/pub/unix/database/Onyx/Onyx.2.45.src.tar.gz">ftp://ftp.uni-bremen.de/pub/unix/database/Onyx/Onyx.2.45.src.tar.gz</a>
    (195872 bytes)
</pre>
 
<dt>(=) <b><a name="oscafe">Open Systems Cafe</a></b>
 
    <dd> Rich Bryant <a href="mailto:rbb@iseek.com">
    &lt;rbb@iseek.com&gt;</a> and Julian Pickard
    <a href="mailto:julian@iseek.com"> &lt;julian@iseek.com&gt;</a>
    have established the Open Systems Cafe. &quot;The Cafe is designed
    to be an up to date information clearing house for anything
    associated with open systems computing&quot;. The Cafe has a
    discussion area dedicated to mSQL. For more details see:
 
<pre>
    <a href="http://iseek.com/Cafe">http://iseek.com/Cafe</a>
</pre>
 
<dt><b><a name="os2">OS/2</a></b>
 
    <dd> Dirk Ohme
    <a href="mailto:dohme@transtec.de">
    &lt;dohme@transtec.de&gt;</a> has ported mSQL 1.0.16 to OS/2. This
    version now includes an ODBC-compliant manager/driver within the
    package. It is available via anonymous ftp from:  <p>
 
<pre>
    <a href="ftp://ftp.nerosworld.com/pub/msql/contrib/mSQLOS2/msql116e.lsm">ftp://ftp.nerosworld.com/pub/msql/contrib/mSQLOS2/msql116e.lsm</a>
    (1913 bytes)
 
    <a href="ftp://ftp.nerosworld.com/pub/msql/contrib/mSQLOS2/msql116e.zip">ftp://ftp.nerosworld.com/pub/msql/contrib/mSQLOS2/msql116e.zip</a>
    (871729 bytes)
</pre>
 
    Dirk has also finished a beta version of mSQL 2.0 for OS/2. This
    release is based on mSQL 2.0B6. It is available via anonymous ftp
    from: <p>
 
<pre>
    <a href="ftp://ftp.nerosworld.com/pub/msql/contrib/mSQLOS2/msql20b6.lsm">ftp://ftp.nerosworld.com/pub/msql/contrib/mSQLOS2/msql20b6.lsm</a>
    (2228 bytes)
 
    <a href="ftp://ftp.nerosworld.com/pub/msql/contrib/mSQLOS2/msql20b6.zip">ftp://ftp.nerosworld.com/pub/msql/contrib/mSQLOS2/msql20b6.zip</a>
    (1865128 bytes)
</pre>
 
    This version also requires Dirk's intrinsic ODBC package. For
    details regarding this package see the
    <a href="#iODBC">&quot;ODBC&quot;</a> section above. <p>
 
<dt><b><a name="perl">Perl</a></b>
 
    <dd> Andreas Koenig
    <a href="mailto:a.koenig@mind.de">&lt;a.koenig@mind.de&gt;</a> has
    contributed a Perl 5 module which allows perl to interface to mSQL
    databases. The latest version is available via anonymous ftp from
    any of the CPAN archives, for example:  <p>
 
<pre>
    <a href="ftp://ftp.funet.fi/pub/languages/perl/CPAN/modules/by-category/07_Database_Interfaces/Msql">ftp://ftp.funet.fi/pub/languages/perl/CPAN/modules/by-category/07_Database_Interfaces/Msql</a>
</pre>
 
    For more information about CPAN see: <p>
 
<pre>
    <a href="ftp://ftp.funet.fi/pub/languages/perl/CPAN">ftp://ftp.funet.fi/pub/languages/perl/CPAN</a>
</pre>
 
    There is a mailing list for MsqlPerl. Send postings to
    <a href="mailto:msqlperl@franz.ww.tu-berlin.de">
    &lt;msqlperl@franz.ww.tu-berlin.de&gt;</a>. <p>
 
<dt><b><a name="php">PHP - Home Page Construction Kit</a></b>
 
    <dd> Rasmus Lerdorf
    <a href="mailto:rasmus@vex.net">&lt;rasmus@vex.net&gt;</a> has
    developed a package that allows users to create WWW sites with mSQL
    databases. <p>
 
    Rasmus writes: <p>
 
<pre>
    PHP/FI is a server-side html-embedded scripting language with
    built-in access logging, access restriction, as well as support for
    ndbm, gdbm and mSQL databases. It also implements the RFC-1867
    standard for form-based file uploads.
 
    The mSQL support is just a small set of functions supported by the
    package. A full set of string manipulation, regular expression,
    directory and file routines complement the script language.
</pre>
 
    The source distribution as well as more information is available
    at: <p>
 
<pre>
    <a href="http://www.vex.net/php">http://www.vex.net/php</a>
</pre>
 
<dt><b><a name="pts">PTS</a></b>
 
    <dd> Dave Shevett <a href="mailto:shevett@homeport.org">
    &lt;shevett@homeport.org&gt;</a> has developed a web based project
    tracking system - PTS. It used mSQL and PHP/FI. For more details
    see: <p>
 
<pre>
    <a href="http://www.homeport.org/~shevett/pts">http://www.homeport.org/~shevett/pts</a>
</pre>
 
<dt><b><a name="python">Python</a></b>
 
    <dd> Anthony Baxter
    <a href="mailto:anthony.baxter@aaii.oz.au">
    &lt;anthony.baxter@aaii.oz.au&gt;</a> has provided an mSQL
    extension to the Python language. It is available via anonymous ftp
    from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLPython/PymSQL.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLPython/PymSQL.tar.gz</a>
    (7581 bytes)
</pre>
 
    Mark Shuttleworth
    <a href="mailto:marks@aztec.co.za">
    &lt;marks@aztec.co.za&gt;</a> has improved upon Anthony's original
    work. Mark's work is available via anonymous ftp from: <p>
 
<pre>
    <a href="http://www.python.org/ftp/python/contrib/Database/mSQL.tar.gz">http://www.python.org/ftp/python/contrib/Database/mSQL.tar.gz</a>
    (6000 bytes)
</pre>
 
    Jeffrey Ollie <a href="mailto:jeff@ollie.clive.ia.us">
    &lt;jeff@ollie.clive.ia.us&gt;</a> has updated the Python module
    for use with both mSQL version 1.<i>x</i> and 2.<i>x</i>. More
    details can be found at: <p>
 
<pre>
    <a href="http://www.ollie.clive.ia.us/python/msql">http://www.ollie.clive.ia.us/python/msql</a>
</pre>
 
<dt><b><a name="radius">RADIUS</a></b>
 
    <dd> Tracy Snell <a href="mailto:waz@enteract.com">
    &lt;waz@enteract.com&gt;</a> has modified
    <a href="http://www.livingston.com">Livingston's</a> RADIUS
    security server to place accounting records in an mSQL database.
    For more details see: <p>
 
<pre>
    <a href="http://www.enteract.com/isp">http://www.enteract.com/isp</a>
</pre>
 
    According to Damian Hamill <a href="mailto:damian@cablenet.net">
    &lt;damian@cablenet.net&gt;</a>, the RADIUS server bundled with
    <a href="http://www.cablenet.net"> Cablenet's</a>
    <a href="http://www.cablenet.net/cablenet/popinabox">PoP In A
    Box</a> ISP server software also supports mSQL based accounting
    records. For more details see: <p>
 
<pre>
    <a href="http://www.cablenet.net/cablenet/popinabox">http://www.cablenet.net/cablenet/popinabox</a>
</pre>
 
<dt><b><a name="rexx">REXX</a></b>
 
    <dd> Mark Hessling
    <a href="mailto:m.hessling@qut.edu.au">
    &lt;m.hessling@qut.edu.au&gt;</a> has provided an mSQL extension to
    REXX under Un*x and <a href="#os2">OS/2</a>. It is available via
    anonymous ftp from the following sites: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLREXX">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLREXX</a>
    <a href="ftp://ftp.qut.edu.au/src/REXXSQL">ftp://ftp.qut.edu.au/src/REXXSQL</a>
    <a href="ftp://ftp.xylogics.com/pub/misc/REXXSQL">ftp://ftp.xylogics.com/pub/misc/REXXSQL</a>
</pre>
 
    You'll need to download the following files: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLREXX/rxsqldoc13.lsm">rxsqldoc13.lsm</a> (1014 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLREXX/rxsqldoc13.zip">rxsqldoc13.zip</a> (39170 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLREXX/rxsqlmin13.lsm">rxsqlmin13.lsm</a> (1001 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLREXX/rxsqlmin13.zip">rxsqlmin13.zip</a> (156525 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLREXX/rxsqlsam13.lsm">rxsqlsam13.lsm</a> (1014 bytes)
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLREXX/rxsqlsam13.zip">rxsqlsam13.zip</a> (25299 bytes)
</pre>
 
<dt><b><a name="simple-sql">Simple SQL</a></b>
 
    <dd> Brian Jepson
    <a href="mailto:bjepson@janus.saturn.net">
    &lt;bjepson@janus.saturn.net&gt;</a> has developed a fairly
    full-featured database management system toolkit using
    <a href="#php">PHP</a> and mSQL. It is available from: <p>
 
<pre>
    <a href="http://www.saturn.net/~bjepson/simple.html">http://www.saturn.net/~bjepson/simple.html</a>
</pre>
 
    Brian has begun work on the successor to Simple SQL. Msql-RDBMS
    &quot;will eventually be a complete relational database management
    system for Msql. It uses HTML forms as an interface&quot;. It is
    available from: <p>
 
<pre>
    <a href="http://www.perl.com/CPAN/modules/by-module/Msql">http://www.perl.com/CPAN/modules/by-module/Msql</a>
</pre>
 
<dt><b><a name="sqlbase">SQLBase</a></b>
 
    <dd> Klaus Thiele
    <a href="mailto:kth@oblib.teuto.de">&lt;kth@oblib.teuto.de&gt;</a>
    has provided a wrapper to allow SQLBase users to interface to mSQL
    databases. It is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLBase/mSQLBase-1.00.tgz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLBase/mSQLBase-1.00.tgz</a>
</pre>
 
<dt><b><a name="tcl">Tcl</a></b>
 
    <dd> Hakan Soderstrom
    <a href="mailto:hs@soderstrom.se">&lt;hs@soderstrom.se&gt;</a> has
    provided an mSQL extension to the Tcl language. It has been tested
    with Tcl 7.5, Tk 4.1 and mSQL 1.0.16 under SunOS 4.1.4. Successful
    ports to several other platforms have been reported. It is
    available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLTCL/msqltcl-1.99.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLTCL/msqltcl-1.99.tar.gz</a>
    (69356 bytes)
</pre>
 
    <dd> Brad Pepers
    <a href="mailto:pepersb@cuug.ab.ca">&lt;pepersb@cuug.ab.ca&gt;</a>
    has also provided an mSQL extension to Tcl. According to the
    documentation it supports tcl7.3 and tk3.6. It is available via
    anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLTCL/tcl_msql.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLTCL/tcl_msql.tar.gz</a>
    (7998 bytes)
</pre>
 
<dt><b><a name="time-utils">Time and date utilities</a></b>
 
    <dd> Pascal Forget
    <a href="mailto:pascal@wsc.com">&lt;pascal@wsc.com&gt;</a> has
    contributed a library of time and date conversion utilities. It is
    available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLtime/time_library.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLtime/time_library.tar.gz</a>
    (7989 bytes)
</pre>
 
<dt><b><a name="tkmsql">tkmSQL</a></b>
 
    <dd> Alligator Decartes
    <a href="mailto:descarte@hermetica.com">
    &lt;descarte@hermetica.com&gt;</a> is developing an mSQL interface
    that uses Perl 5, Tk, and DBD/DBD-mSQL. It is currently in alpha
    release for developers only but Alligator would appreciate
    <a href="mailto:descarte@hermetica.com">feedback</a>. <p>
 
    tkmSQL requires: <p>
 
<ul>
    <li> perl 5.001m or above
    <li> Tk-b8 <p>
 
    <b>Note:</b> This is a perl Tk module and is NOT to be confused
    with Tk itself. It can be obtained via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.wpi.edu/perl5">ftp://ftp.wpi.edu/perl5</a>
</pre>
 
    There is also a FAQ available from: <p>
 
<pre>
    <a href="http://w4.lns.cornell.edu/~pvhp/ptk/ptkFAQ.html">http://w4.lns.cornell.edu/~pvhp/ptk/ptkFAQ.html</a>
</pre>
 
    <li> <a href="#dbi">DBI-0.65</a>
    <li> <a href="#dbd-msql">DBD::mSQL-0.60pl9</a>
</ul>
<p>
 
    You may obtain tkmSQL via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.mcqueen.com/pub/databases/dbatools/tkmSQL">ftp://ftp.mcqueen.com/pub/databases/dbatools/tkmSQL</a>
</pre>
 
<dt><b><a name="sgs">Unique sequence number generator</a></b>
 
    <dd> Pascal Forget
    <a href="mailto:pascal@wsc.com">&lt;pascal@wsc.com&gt;</a> has
    contributed a unique sequence number generator that can be used by
    mSQL applications (and others) to provide unique identifiers. It is
    available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLsgs/sgs-1.0.0.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLsgs/sgs-1.0.0.tar.gz</a>
    (24216 bytes)
</pre>
 
<dt><b><a name="vdb">Virtual Database</a></b>
 
    <dd> Damian Hamill <a href="mailto:damian@cablenet.net">
    &lt;damian@cablenet.net&gt;</a> has written a virtual database
    layer that sits between your application and mSQL. This provides a
    level of abstraction between your application and the database
    back end such that it becomes a fairly simple matter to change
    database systems without changing your application code. The
    current version is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLVdb/Vdb.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLVdb/Vdb.tar.gz</a>
    (61670 bytes)
</pre>
 
    Damian is working on the next release which will include support
    for a number of different database engines such as mSQL, mysql,
    Sybase and Informix. <p>
 
<dt><b><a name="virtuflex">VirtuFlex</a></b>
 
    <dd> VirtuFlex Software has added mSQL support to their web/database
    development tool VirtuFlex. More details can be seen at: <p>
 
<pre>
    <a href="http://www.virtuflex.com">http://www.virtuflex.com</a>
</pre>
 
<dt><b><a name="visual-basic">Visual Basic</a></b>
 
    <dd> Jim Gerace
    <a href="mailto:kasi@dreamscape.com">
    &lt;kasi@dreamscape.com&gt;</a> has implemented a
    Visual Basic client API for MS Windows and mSQL. It is available
    via anonymous ftp from:
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLVB/msqlcvb.zip">ftp://bond.edu.au/pub/Minerva/msql/Contrib/mSQLVB/msqlcvb.zip</a>
    (42192 bytes)
</pre>
 
<dt><b><a name="wdb">WDB</a></b>
 
    <dd> Bo Frese Rasmussen
    <a href="mailto:bfr@dtv.dk">&lt;bfr@dtv.dk&gt;</a> has
    developed a Web database interface called WDB. <p>
 
    WDB is a software tool set that tremendously simplifies the
    integration of SQL based databases into the World Wide Web. WDB
    lets you provide WWW access to the contents of databases without
    writing a single line of code! <p>
 
    At the moment WDB supports Sybase, Informx and mSQL. However it is
    relatively easy to port it to other SQL based databases. <p>
 
    For more details on WDB see: <p>
 
<pre>
    <a href="http://www.dtv.dk/~bfr/wdb">http://www.dtv.dk/~bfr/wdb</a>
</pre>
 
    Jeff Rowe <a href="mailto:beowulf@cscsun4.larc.nasa.gov">
    &lt;beowulf@cscsun4.larc.nasa.gov&gt;</a> has published a tutorial
    on enhancing WDB. Details can be found at: <p>
 
<pre>
    <a href="http://cscsun1.larc.nasa.gov/~beowulf/tutor/index.html">http://cscsun1.larc.nasa.gov/~beowulf/tutor/index.html</a>
</pre>
 
<dt><b><a name="websql">Websql</a></b>
 
    <dd> Henry Minsky
    <a href="mailto:hqm@ua.com">&lt;hqm@ua.com&gt;</a> has developed an
    mSQL Table WWW browser interface. <p>
 
    &quot;This is a C web CGI script to examine and modify rows in
    tables of an mSQL database. You should use Netscape or another
    browser which supports HTML 3.0 tables.&quot; <p>
 
    More details and sample output are available from: <p>
 
<pre>
    <a href="http://www.ua.com/websql">http://www.ua.com/websql</a>
</pre>
 
    Source code is available from: <p>
 
<pre>
    <a href="http://www.ua.com/websql/websql.tar.gz">http://www.ua.com/websql/websql.tar.gz</a>
    (24225 bytes)
</pre>
 
<dt><b><a name="tryc-repository">Wojciech Tryc's Repository</a></b>
 
    <dd> Wojciech Tryc <a href="mailto:wojtek@tryc.on.ca">
    &lt;wojtek@tryc.on.ca&gt;</a> has established a repository of mSQL
    and PHP contributed software. It can be found at: <p>
 
<pre>
    <a href="http://solaris.tryc.on.ca/files/files.phtml">http://solaris.tryc.on.ca/files/files.phtml</a>
</pre>
 
<dt><b><a name="xfsql">XfSQL</a></b>
 
    <dd> Mark Loveland
    <a href="mailto:mark@zeus.mysticgrp.com">
    &lt;mark@zeus.mysticgrp.com&gt;</a> has developed an X interface to
    mSQL using the Xforms package. Mark's work is available via
    anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/XfSQL/xfsql.tar.gz">ftp://bond.edu.au/pub/Minerva/msql/Contrib/XfSQL/xfsql.tar.gz</a>
    (74596 bytes)
</pre>
 
    Xforms is available for a number of different platforms via
    anonymous ftp from either of the following locations: <p>
 
<pre>
    <a href="ftp://einstein.phys.uwm.edu/pub/xforms">ftp://einstein.phys.uwm.edu/pub/xforms</a>
</pre>
 
<dt><b><a name="xsqlmenu">Xsqlmenu</a></b>
 
    <dd> Kees Lemmens
    <a href="mailto:lemmens@dv.twi.tudelft.nl">
    &lt;lemmens@dv.twi.tudelft.nl&gt;</a> has also developed an X
    interface to mSQL using the Xforms package. Xsqlmenu can be
    obtained via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ta.twi.tudelft.nl/pub/dv/lemmens/xsqlmenu_1.02s.tar.gz">ftp://ta.twi.tudelft.nl/pub/dv/lemmens/xsqlmenu_1.02s.tar.gz</a>
    (14973 bytes)
</pre>
 
    A pre compiled binary for Linux is also available via anonymous ftp
    from: <p>
 
<pre>
    <a href="ftp://ta.twi.tudelft.nl/pub/dv/lemmens/xsqlmenu_1.02LinuxBin.tar.gz">ftp://ta.twi.tudelft.nl/pub/dv/lemmens/xsqlmenu_1.02LinuxBin.tar.gz</a>
    (130945 bytes)
</pre>
 
    Xforms is available for a number of different platforms via
    anonymous ftp from either of the following locations: <p>
 
<pre>
    <a href="ftp://einstein.phys.uwm.edu/pub/xforms">ftp://einstein.phys.uwm.edu/pub/xforms</a>
</pre>
 
    Kees has upgraded Xsqlmenu to support mSQL 2.<i>x</i>. Source code
    is available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ta.twi.tudelft.nl/pub/dv/lemmens/xsqlmenu.2.01b.tar.gz">ftp://ta.twi.tudelft.nl/pub/dv/lemmens/xsqlmenu.2.01b.tar.gz</a>
    (18601 bytes)
</pre>
 
<dt><b><a name="cplusplus">Z Classes for C++</a></b>
 
    <dd> Dean Fuqua
    <a href="mailto:fuqua@niehs.nih.gov">
    &lt;fuqua@niehs.nih.gov&gt;</a> has also contributed a set of C++
    classes to provide access to both Oracle and mSQL databases. It is
    available via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://bond.edu.au/pub/Minerva/msql/Contrib/zmSQL/zmsql-2.1.tar">ftp://bond.edu.au/pub/Minerva/msql/Contrib/zmSQL/zmsql-2.1.tar</a>
    (40960 bytes)
</pre>
</dl>
 
<hr>
 
<center>
<h2><a name="web-sites">Web sites using mSQL</a></h2>
</center>
 
The following web sites are using mSQL as the database component of their
pages. <p>
 
<center>
<table border=3 cellpadding=3 bgcolor="#cccccc">
    <tr>
	<td>
	    <a href="#3d-planet">3D Planet</a>
	</td>
 
	<td>
	    <a href="#isp">ISP Listing</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#allens-flowers">Allen's Flowers</a>
	</td>
 
	<td>
	    <a href="#kelloggs">Kelloggs</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#aussie-index">Aussie Index</a>
	</td>
 
	<td>
	    <a href="#kids-health">Kids Health</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#bishop-museum">Bishop Museum, Honolulu</a>
	</td>
 
	<td>
	    <a href="#mahrens-addressbook">Matthew Ahrens' Address Book</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#bond-staff">Bond University Staff Index</a>
	</td>
 
	<td>
	    <a href="#mondo-db">Mondo-DB Project</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#borsen">Borsen</a>
	</td>
 
	<td>
	    <a href="#nerosworld">Nero's World</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#canadait">CanadaIT</a>
	</td>
 
	<td>
	    <a href="#netcal">NetCal!</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#cinemabrazil">Cinema Brazil</a>
	</td>
 
	<td>
	    <a href="#newshare">Newshare Corporation</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#consecol">Conservation Ecology</a>
	</td>
 
	<td>
	    <a href="#ncid">North Carolina Insurance Database</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#cyberclip">CyberClip</a>
	</td>
 
	<td>
	    <a href="#opc">Ontario Prevention Clearinghouse</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#dmx-cables">DMX Cable Systems</a>
	</td>
 
	<td>
	    <a href="#paulinasprings">Paulina Springs Book Company</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#dave-gizmo">Dave 'Gizmo' Gymer</a>
	</td>
 
	<td>
	    <a href="#physics">Physics Pilot</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#europractice">EUROPRACTICE</a>
	</td>
 
	<td>
	    <a href="#qms">QMS</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#ewse-org">European Wide Service Exchange</a>
	</td>
 
	<td>
	    <a href="#querri">QUERRI Bibliographic Server</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#first-byte">First Byte Data Systems</a>
	</td>
 
	<td>
	    <a href="#railway">The Railway Exchange</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#gcrc">Government Contractor Resource Center</a>
	</td>
 
	<td>
	    <a href="#used-gear">Used gear price list</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    (=) <a href="#iecafe">Internet Employment Cafe</a>
	</td>
 
	<td>
	    <a href="#web-ventures">Web Ventures</a>
	</td>
    </tr>
 
    <tr>
	<td>
	    <a href="#iq-now">The IQ NOW Member Directory</a>
	</td>
 
	<td>
	    <a href="#wise">Web Integrated Software metrics Environment</a>
	</td>
    </tr>
</table>
</center>
<p>
 
<dl>
 
<dt><b><a name="3d-planet">3D Planet</a></b>
 
    <dd> Mark Mazur
    <a href="mark@opencad.com">&lt;mark@opencad.com&gt;</a>
    uses mSQL and <a href="#php">The Personal Home Page Construction
    Kit </a> to enable users to find the nearest 3D Planet store. This
    page can be seen at: <p>
 
<pre>
    <a href="http://www.3dplanet.com/d5.html">http://www.3dplanet.com/d5.html</a>
</pre>
 
<dt><b><a name="allens-flowers">Allen's Flowers</a></b>
 
    <dd> As you might expect, Allen's Flowers sells flowers. They use
    mSQL to drive their ordering system. Their web site can be seen
    at: <p>
 
<pre>
    <a href="http://www.allensflowers.com">http://www.allensflowers.com</a>
</pre>
 
<dt><b><a name="aussie-index">Aussie Index</a></b>
 
    <dd> Aussie.com.au uses mSQL to collate its list of Australian Web
    sites. Their web page can be seen at: <p>
 
<pre>
    <a href="http://www.aussie.com.au">http://www.aussie.com.au</a>
</pre>
 
<dt><b><a name="bishop-museum">Bishop Museum, Honolulu</a></b>
 
    <dd> Jeffrey Sue
    <a href="mailto:jysue@aloha.net">&lt;jysue@aloha.net&gt;</a>
    has used mSQL and Rasmus Lerdorf's <a href="#php">PHP</a>
    package to create the first of several web based
    biology related databases. The first database is the arthropod
    checklist, a list of all known arthropods found in the Hawaiian
    Islands. The
    next databases will be additional checklists, starting with
    flowering
    plants. It can be seen at: <p>
 
<pre>
    <a href="http://www.bishop.hawaii.org/bishop/HBS/arthrosearch.html">http://www.bishop.hawaii.org/bishop/HBS/arthrosearch.html</a>
</pre>
 
<dt><b><a name="bond-staff">Bond University Staff Index</a></b>
 
    <dd> The staff index at Bond University, Queensland Australia uses
    mSQL to power its staff address database. See their web page for
    more details: <p>
 
<pre>
    <a href="http://www.Bond.edu.au/Bond/General/MParse/mparse.html">http://www.Bond.edu.au/Bond/General/MParse/mparse.html</a>
</pre>
 
<dt><b><a name="borsen">Borsen</a></b>
 
    <dd> This Danish site uses mSQL as its databases for searches. I
    can't give more information as it is all in Danish and my Danish
    is truly awful. For more details visit: <p>
 
<pre>
    <a href="http://www.borsen.dk">http://www.borsen.dk</a>
</pre>
 
<dt><b><a name="canadait">CanadaIT</a></b>
 
    <dd> CanadaIT limited uses mSQL. Their web page can be seen at:
    <p>
 
<pre>
    <a href="http://www.CanadaIT.com"> http://www.CanadaIT.com</a>
</pre>
 
<dt><b><a name="cinemabrazil">Cinema Brazil</a></b>
 
    <dd> Marcos Manhaes Marins
    <a href="mailto:cinemabrazil@ax.apc.org">
    &lt;cinemabrazil@ax.apc.org&gt;</a> has used mSQL to power the
    search engine behind the Cinema Brazil web site. For more details
    see: <p>
 
<pre>
    <a href="http://www.ibase.org.br/~cinemabrazil">http://www.ibase.org.br/~cinemabrazil</a>
</pre>
 
<dt><b><a name="consecol">Conservation Ecology Journal</a></b>
 
    <dd> Darryl Staflund <a href="dstaflun@ccs.carleton.ca">
    &lt;dstaflun@ccs.carleton.ca&gt;</a> uses mSQL behind the online
    journal Conservation Ecology. It can be seen at: <p>
 
<pre>
    <a href="http://www.consecol.org">http://www.consecol.org</a>
</pre>
 
<dt><b><a name="cyberclip">CyberClip</a></b>
 
    <dd> CyberClip uses mSQL to power its Shopper's Index search engine.
    Their web page can be seen at: <p>
 
<pre>
    <a href="http://www.cyberclip.com">http://www.cyberclip.com</a>
</pre>
 
<dt><b><a name="dmx-cables">DMX Cables</a></b>
 
    <dd> Digital Music Express uses mSQL to allow users to search for
    DMX cable service providers in their area. Their web site can be
    seen at: <p>
 
<pre>
    <a href="http://www.dmxmusic.com">http://www.dmxmusic.com</a>
</pre>
 
<dt><b><a name="dave-gizmo">Dave 'Gizmo' Gymer</a></b>
 
    <dd> Dave Gymer
    <a href="mailto:dgymer@gdcarc.co.uk">
    &lt;dgymer@gdcarc.co.uk&gt;</a> has used mSQL and Dean Fuqua's
    <a href="#cplusplus">zclass</a> C++ classes to provide a web based
    database system for his collection of compact discs. The CD
    collection and the source code can be seen at: <p>
 
<pre>
    <a href="http://www.mal.com/~dgymer/gizmo/music.html">http://www.mal.com/~dgymer/gizmo/music.html</a>
</pre>
 
<dt><b><a name="europractice">EUROPRACTICE</a></b>
 
    <dd> EUROPRACTICE is an ASIC manufacturing service. Their home
    page can be found at: <p>
 
<pre>
    <a href="http://www.imec.be/europractice/europractice.html">http://www.imec.be/europractice/europractice.html</a>
</pre>
 
 
<dt><b><a name="ewse-org">European Wide Service Exchange</a></b>
 
    <dd> The <a href="http://ewse.ceo.org">European Wide Service
    Exchange</a> uses mSQL extensively. Apart from a user registration
    and customization service, complex relational tables allow context
    sensitive searching, both by geographic area as well as by 'free
    text'. Further functionality relying on mSQL includes urn-&gt;urc
    resolution services, calendars and an automatic 'what is new page'.
    What makes this site unique however is the ability for the user to
    add, modify or remove data entries from the databases through a web
    interface. <p>
 
    Their home page can be found at: <p>
 
<pre>
    <a href="http://ewse.ceo.org">http://ewse.ceo.org</a>
</pre>
 
<dt><b><a name="first-byte">First Byte Data Systems</a></b>
 
    <dd> First Byte Data Systems uses mSQL to power its product search
    engine. Their home page can be seen at: <p>
 
<pre>
    <a href="http://www.first-byte.com">http://www.first-byte.com</a>
</pre>
 
<dt><b><a name="gcrc">Government Contractor Resource Center</a></b>
 
    <dd> The Government Contractor Resource Center provides
    information for companies wishing to provide contractual services
    to United States Government bodies. Their web site can be seen at:
    <p>
 
<pre>
    <a href="http://www.govcon.com">http://www.govcon.com</a>
</pre>
 
<dt>(=) <b><a name="iecafe">Internet Employment Cafe</a></b>
 
    <dd> The Employment Services Department of the University of
    Western Ontario has established an Internet Employment Cafe
    powered by mSQL. It can be seen at: <p>
 
<pre>
    <a href="http://cafe.sdc.uwo.ca">http://cafe.sdc.uwo.ca</a>
</pre>
 
<dt><b><a name="iq-now">The IQ Now Member Directory</a></b>
 
    <dd> IQ Now provides information on healthcare services. Their web
    site is at: <p>
 
<pre>
    <a href="http://www.iqnow.com">http://www.iqnow.com</a>
</pre>
 
<dt><b><a name="isp">ISP Listing</a></b>
 
    <dd> Rasmus Lerdorf
    <a href="mailto:rasmus@vex.net">&lt;rasmus@vex.net&gt;</a>
    has used mSQL and PHP/FI to maintain a list of Internet Service
    Providers. His work can be seen at: <p>
 
<pre>
    <a href="http://www.vex.net/isp">http://www.vex.net/isp</a>
</pre>
 
<dt><b><a name="kelloggs">Kelloggs</a></b>
 
    <dd> Kelloggs (the Corn Flakes people) use mSQL. Their web site can
    be seen at: <p>
 
<pre>
    <a href="http://www.kelloggs.com">http://www.kelloggs.com</a>
</pre>
 
<dt><b><a name="kids-health">Kids Health</a></b>
 
    <dd> KidsHealth is a resource centre for matters relating to
    children's health. See their home page at: <p>
 
<pre>
    <a href="http://KidsHealth.org">http://KidsHealth.org</a>
</pre>
 
<dt> <b><a name="mahrens-addressbook">Matthew Ahrens' Address Book</a></b>
 
    <dd> Matthew Ahrens
    <a href="mailto:matt@callnet.com">
    &lt;matt@callnet.com&gt;</a> has written a Web based address book.
    It uses Perl and mSQL. It can be found at: <p>
 
<pre>
    <a href="http://www.callnet.com/~matt">http://www.callnet.com/~matt</a>
</pre>
 
<dt><b><a name="mondo-db">Mondo-DB Project</a></b>
 
    <dd> Rose-Hulman Institute of Technology is using mSQL to manage
    much of its Unix networking data. Their work can be seen at: <p>
 
<pre>
    <a href="http://www.rose-hulman.edu/~allard/Mondo-DB/oindex.html">http://www.rose-hulman.edu/~allard/Mondo-DB/oindex.html</a>
</pre>
 
<dt><b><a name="nerosworld">Nero's World</a></b>
 
    <dd> David Perry <a href="mailto:deperry@nerosworld.com">
    &lt;deperry@nerosworld.com&gt;</a> uses mSQL extensively to drive
    his web server. Examples of his work can be seen at:  <p>
 
<pre>
    <a href="http://www.nerosworld.com/realestate/">http://www.nerosworld.com/realestate/</a>
    <a href="http://www.nerosworld.com/business/">http://www.nerosworld.com/business/</a>
    <a href="http://www.nerosworld.com/tradingpost/">http://www.nerosworld.com/tradingpost/</a>
    <a href="http://www.nerosworld.com/fstop/">http://www.nerosworld.com/fstop/</a>
    <a href="http://www.nerosworld.com/nero/zipcode.htm">http://www.nerosworld.com/nero/zipcode.htm</a>
    <a href="http://www.nerosworld.com/romancing_the_web/">http://www.nerosworld.com/romancing_the_web/</a>
 
</pre>
 
<dt><b><a name="netcal">NetCal!</a></b>
 
    <dd> NetCal! is an interactive calendar system. It can be found
    at: <p>
 
<pre>
    <a href="http://www.itribe.net/netcal">http://www.itribe.net/netcal</a>
</pre>
 
<dt><b><a name="newshare">Newshare Corporation</a></b>
 
    <dd> Newshare Corporation is an internet news broker. Their home
    page can be seen at: <p>
 
<pre>
    <a href="http://www.newshare.com">http://www.newshare.com</a>
</pre>
 
<dt><b><a name="ncid">North Carolina Insurance Database</a></b>
 
    <dd> North Carolina Insurance Database provides a registry of
    insurance providers in North Carolina. They can be found at: <p>
 
<pre>
    <a href="http://www.nc-insurance.com">http://www.nc-insurance.com</a>
</pre>
 
<dt><b><a name="opc">Ontario Prevention Clearinghouse</a></b>
 
    <dd> Ontario Prevention Clearinghouse provides a resource centre
    for community based programs. Their home page can be found at: <p>
 
<pre>
    <a href="http://www.opc.on.ca">http://www.opc.on.ca</a>
</pre>
 
<dt><b><a name="paulinasprings">Paulina Springs Book Company</a></b>
 
    <dd> The Paulina Springs Book Company uses mSQL and PHP/FI to help
    drive it's web site. Visit it at: <p>
 
<pre>
    <a href="http://www.paulinasprings.com">http://www.paulinasprings.com</a>
</pre>
 
<dt><b><a name="physics">Physics Pilot</a></b>
 
    <dd> Kenneth Holmlund
    <a href="mailto:Kenneth.Holmlund@TP.UmU.SE">
    &lt;Kenneth.Holmlund@TP.UmU.SE&gt;</a> from the Department of
    Theoretical Physics at Umee University in Sweden has used mSQL and
    <a href="#php">The Personal Home Page Construction Kit</a> to
    create the &quot;The Internet Pilot to Physics&quot;. It can be
    seen at: <p>
 
<pre>
    <a href="http://www.tp.umu.se/TIPTOP">http://www.tp.umu.se/TIPTOP</a>
</pre>
 
<dt><b><a name="qms">QMS</a></b>
 
    <dd> James Hill
    <a href="mailto:james_hill@iscclink.is.qms.com">
    &lt;james_hill@iscclink.is.qms.com&gt;</a> has used mSQL and
    <a href="#php">The Personal Home Page Construction Kit</a> to
    create a searchable web based FAQ for QMS printers. It can be seen
    at: <p>
 
<pre>
    <a href="http://www.qms.com/www/faq">http://www.qms.com/www/faq</a>
</pre>
 
    James has made the source code available via anonymous ftp. It can
    be downloaded from: <p>
 
<pre>
    <a href="ftp://ftp.qms.com/pub/mktg/outgoing/SupportBase.tar.gz">ftp://ftp.qms.com/pub/mktg/outgoing/SupportBase.tar.gz</a>
    (5509 bytes)
</pre>
 
<dt><b><a name="querri">QUERRI Bibliographic Server</a></b>
 
    <dd> QUERRI or &quot;Questions on University Regional Resource
    Information&quot; is an online database with bibliographic
    information on more than 15,000 educational resources within the
    United States. It can be accessed at: <p>
 
<pre>
    <a href="http://idea.exnet.iastate.edu:8080">http://idea.exnet.iastate.edu:8080</a>
</pre>
 
<dt><b><a name="railway">The Railway Exchange</a></b>
 
    <dd> Lester Hightower
    <a href="hightowe@scri.fsu.edu">&lt;hightowe@scri.fsu.edu&gt;</a>
    uses mSQL to power his web pages on model railway equipment. They
    can be seen at: <p>
 
<pre>
    <a href="http://www.railwayex.com/">http://www.railwayex.com/</a>
</pre>
 
<dt><b><a name="used-gear">Used gear price list</a></b>
 
    <dd> Neil Bradley
    <a href="mailto:neil@synthcom.com">&lt;neil@synthcom.com&gt;</a>
    has used mSQL to provide a price list for second hand music
    equipment. A demonstration of this system is available at <p>
 
<pre>
    <a href="http://www.synthcom.com/cgi-bin/gear">http://www.synthcom.com/cgi-bin/gear</a>
</pre>
 
    and the source code can be obtained via anonymous ftp from: <p>
 
<pre>
    <a href="ftp://ftp.synthcom.com/pub/stuff">ftp://ftp.synthcom.com/pub/stuff</a>
</pre>
 
<dt><b><a name="virtubank">Virtubank</a></b>
 
    <dd>
 
<pre>
    <a href="http://w3.e-sense.net/e-sense/Experimental/Virtu">http://w3.e-sense.net/e-sense/Experimental/Virtu</a>
</pre>
 
<dt><b><a name="web-ventures">Web Ventures</a></b>
 
    <dd> Web Ventures provides WWW services to business. Their home
    page can be seen at: <p>
 
<pre>
    <a href="http://www.webventures.com.au">http://www.webventures.com.au</a>
</pre>
 
<dt><b><a name="wise">Web Integrated Software metrics Environment</a></b>
 
    <dd> The Web Integrated Software metrics Environment (WISE)
    is a project management system with a WWW interface and an mSQL
    back end. Details can be found at: <p>
 
<pre>
    <a href="http://research.ivv.nasa.gov/projects/WISE/wise.html">http://research.ivv.nasa.gov/projects/WISE/wise.html</a>
</pre>
 
</dl>
 
<hr>
<hr>
 
<!-- End of Part 4 -->
 
</body>
</html>