File: notes.html

package info (click to toggle)
erlang-doc-html 1%3A11.b.2-1
links: PTS
area: main
in suites: etch, etch-m68k
size: 23,284 kB
ctags: 10,724
sloc: erlang: 505; ansic: 323; makefile: 62; perl: 61; sh: 45
file content (1727 lines) | stat: -rw-r--r-- 32,670 bytes
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<!-- This document was generated using DocBuilder 3.3.3 -->
<HTML>
<HEAD>
  <TITLE>Inets Release Notes
</TITLE>
  <SCRIPT type="text/javascript" src="../../../../doc/erlresolvelinks.js">
</SCRIPT>
</HEAD>
<BODY BGCOLOR="#FFFFFF" TEXT="#000000" LINK="#0000FF" VLINK="#FF00FF"
      ALINK="#FF0000">
<CENTER>
<A HREF="http://www.erlang.se"><IMG BORDER=0 ALT="[Ericsson AB]" SRC="min_head.gif"></A>
</CENTER>
<A NAME="1"><!-- Empty --></A>
<H2>1 Inets Release Notes
</H2>
<A NAME="1.1"><!-- Empty --></A>
<H3>1.1 Inets 4.7.6</H3>
<A NAME="1.1.1"><!-- Empty --></A>
<H4>1.1.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [httpc] - The parsing of uri's was rewritten so that it
should handle all types of uri's including ipv6 uri's.
<BR>
 Own Id: OTP-5677
<BR>

</LI>


<LI>
  [httpc, httpd] - Extensions and trailers where not
properly handled by the chunk decoding implementation.
<BR>
 Own Id: OTP-6005 Aux Id: OTP-6264 
<BR>

</LI>


<LI>
  [httpc] - A request resulting in an empty body is now
returned without any delays.
<BR>
 Own Id: OTP-6243
<BR>

</LI>


<LI>
  [httpc] - When http:request/4 was used with a configured
proxy, and the webserver returns 3XX code, http:request/4
entered an endless loop. Two problems was solved in this
area, the absolute uri is now updated when a redirect is
issued, so that the problem in this case will not arise,
and the redirection endless loop detection was fixed so
that will actually detect potential endless loops.
<BR>
 Own Id: OTP-6244
<BR>

</LI>


<LI>
  [httpc, httpd] - In some cases if a body contained the
sequence &#34;\r\n0&#34; and was chunked encoded this sequence
was incorrectly interpreted as the last chunk.
<BR>
 Own Id: OTP-6264 Aux Id: OTP-6005 
<BR>

</LI>


<LI>
  [httpc, httpd] - http_request.erl didn't handle https
URIs, which meant that redirects from ESI did not work.
<BR>
 Own Id: OTP-6274
<BR>

</LI>


<LI>
  [httpc, httpd] - The base 64 decoder was missing a guard
so that invalid input lead to an emulator crash instead
of a function clause as expected. Also the http server
has been improved to handle the function clause error
returning a bad credentials reply to the client.
<BR>
 Own Id: OTP-6279
<BR>

</LI>


<LI>
  [httpc] - Changed internal default value as it sometimes
would be interpreted incorrectly causing the client to
return an incomplete body.
<BR>
 Own Id: OTP-6283
<BR>

</LI>


<LI>
  [httpc] - Handling of 30X codes was changed so that it
works according to the documentation. For instance 301
and 302 codes will not be automatically redirected.
<BR>
 Own Id: OTP-6297
<BR>

</LI>


<LI>
  [httpc] - A bug in the pipeline-handling code could cause
a response to be sent to the client with an incorrect
request id.
<BR>
 Own Id: OTP-6303
<BR>

</LI>


</UL>
<A NAME="1.1.2"><!-- Empty --></A>
<H4>1.1.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  [httpc] - Added feature to send header values as they
where typed by the user of the client. Note that the http
standard requires them to be case insensative. This
feature should only be used if there is no other way to
communicate with the server or for testing purpose.
<BR>
 Own Id: OTP-5527
<BR>

</LI>


<LI>
  [httpc] - When using asynchronous HTTP-request it is now
possible to receive &#34;200-responses&#34; as streams instead of
having to wait until the whole response has been
delivered. It also possible to stream &#34;200-response
bodys&#34; to a file both for synchronous and asynchronous
requests.
<BR>
 Own Id: OTP-6263
<BR>

</LI>


<LI>
  [httpc] - Added option to generate Proxy-Authorization
header from provided proxy username and password.
<BR>
 Own Id: OTP-6280
<BR>

</LI>


</UL>
<A NAME="1.2"><!-- Empty --></A>
<H3>1.2 Inets 4.7.5</H3>
<A NAME="1.2.1"><!-- Empty --></A>
<H4>1.2.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [FTP] Change documentation so that it agrees with the
default behaviour of the code regarding the use of
transfer type, which is right according to rfc959.
<BR>
 Own Id: OTP-6018
<BR>

</LI>


<LI>
  [ftp] The application handles the case if the owning
process terminates with the reason 'shutdown'.
<BR>
 Own Id: OTP-6035
<BR>

</LI>


<LI>
  [ftp] Handle file errors from the FTP server.
<BR>
 Own Id: OTP-6036
<BR>

</LI>


<LI>
 [httpd] Header parsing of reply from cgi script incorrect.<BR>

Own Id: OTP-6145<BR>


</LI>


<LI>
 [ftp] The timeout given in the ftp:open call was not
properly used, which could leave the caller hanging
forever.<BR>

Own Id: OTP-6184<BR>

Aux Id: seq10388<BR>


</LI>


<LI>
 [httpd] HTTPD request handler does not handle unexpected info
properly, which causes an unneccessarily obscure error 
message.<BR>

Own Id: OTP-6189<BR>

Aux Id: seq10395<BR>


</LI>


<LI>
 [http] Misc fixes in the URI parsing module.<BR>

Igor Goryachev<BR>

Own Id: OTP-6191<BR>


</LI>


</UL>
<A NAME="1.2.2"><!-- Empty --></A>
<H4>1.2.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
 Added application interface module: <CODE>inets</CODE>.<BR>

Own Id: OTP-6135<BR>


</LI>


</UL>
<A NAME="1.3"><!-- Empty --></A>
<H3>1.3 Inets 4.7.4</H3>
<A NAME="1.3.1"><!-- Empty --></A>
<H4>1.3.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  Removed code generating compiler warnings
<BR>
 Own Id: OTP-6069
<BR>

</LI>


</UL>
<A NAME="1.3.2"><!-- Empty --></A>
<H4>1.3.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
 [tftp] Added documentation (manual page) for 
<A HREF="tftp.html">TFTP</A>.<BR>

Own Id: OTP-6082<BR>


</LI>


</UL>
<A NAME="1.4"><!-- Empty --></A>
<H3>1.4 Inets 4.7.3</H3>
<A NAME="1.4.1"><!-- Empty --></A>
<H4>1.4.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [http,server] started dbg even though no debugging was
desired.
<BR>
 Own Id: OTP-5984 Aux Id: seq10290 
<BR>

</LI>


<LI>
  [http,server] request handler process died ugly due to a
parse error when validating a bad request.
<BR>
 Own Id: OTP-6003 Aux Id: seq10260 
<BR>

</LI>


</UL>
<A NAME="1.5"><!-- Empty --></A>
<H3>1.5 Inets 4.7.2</H3>
<A NAME="1.5.1"><!-- Empty --></A>
<H4>1.5.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [http,server] Server falsely sets ipv6 enabled when it
gets an ipv4-mapped ipv6 address.
<BR>
 Own Id: OTP-5941 Aux Id: seq10221 
<BR>

</LI>


<LI>
  [http,server] In case http version is unknown of client
use HTTP1.0 to send status.
<BR>
 Own Id: OTP-5943 Aux Id: seq10198 
<BR>

</LI>


<LI>
  [http,server] The process handling a request now ignors
garbage messages.
<BR>
 Own Id: OTP-5961 Aux Id: seq10198 
<BR>

</LI>


<LI>
  [http,server] Changed some actions taken if config data
in httpd services was faulty.
<BR>
 Own Id: OTP-5962 Aux Id: seq10198 
<BR>

</LI>


</UL>
<A NAME="1.6"><!-- Empty --></A>
<H3>1.6 inets 4.7.1</H3>
<A NAME="1.6.1"><!-- Empty --></A>
<H4>1.6.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [http,server] It was possible to read arbitrary files on
server by prepending ././ and ../../ in fornt of the file
name.
<BR>

 Own Id: OTP-5938
<BR>


</LI>


</UL>
<A NAME="1.7"><!-- Empty --></A>
<H3>1.7 Inets 4.7</H3>
<A NAME="1.7.1"><!-- Empty --></A>
<H4>1.7.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [http,server] Handling of undefined file times (e.g.
modification time: the mtime field on the file_info
record).
<BR>
 Own Id: OTP-5865 Aux Id: OTP-5844 
<BR>

</LI>


</UL>
<A NAME="1.7.2"><!-- Empty --></A>
<H4>1.7.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  [http,server]It is now possible to set the wanted timeout
for the server to setup a request connection. A new
syntax for inets.config is provided in the users guide
documentation. This syntax also allows to set tracing of
the server for debug purposes.
<BR>
 Own Id: OTP-5913 Aux Id: seq10198 
<BR>

</LI>


</UL>
<A NAME="1.8"><!-- Empty --></A>
<H3>1.8 Inets 4.6.2</H3>
<A NAME="1.8.1"><!-- Empty --></A>
<H4>1.8.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [http,server] Had earlier forgot to convert a value in
entity_body to a binary.
<BR>
 Own Id: OTP-5796
<BR>

</LI>


<LI>
  [http,server] Now application checks whether the
necessary directives under directive Directory exist.
<BR>
 Own Id: OTP-5821
<BR>

</LI>


</UL>
<A NAME="1.9"><!-- Empty --></A>
<H3>1.9 Inets 4.6.1</H3>
<A NAME="1.9.1"><!-- Empty --></A>
<H4>1.9.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [http, client] If an Ipv4-mapped ipv6 address is used the
client now falls back on ipv4.
<BR>
 Own Id: OTP-5773 Aux Id: OTP-5765, OTP-5764 
<BR>

</LI>


<LI>
  [http,server] Content-length may got a too low value,
causing loss of data in clients.
<BR>
 Own Id: OTP-5775 Aux Id: seq10110 
<BR>

</LI>


</UL>
<A NAME="1.9.2"><!-- Empty --></A>
<H4>1.9.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  [http, client] The verbose mode prints what was sent and
received during a request. Added a <CODE>verbose</CODE> option
that can be used by <CODE>http:set_options/1</CODE>
<BR>
 Own Id: OTP-5766
<BR>

</LI>


</UL>
<A NAME="1.10"><!-- Empty --></A>
<H3>1.10 Inets 4.6</H3>
<A NAME="1.10.1"><!-- Empty --></A>
<H4>1.10.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [ftp, client] - Improvement of error handling if
something goes wrong while handling the option list in
ftp:open/[1,2,3].
<BR>
 Own Id: OTP-5711
<BR>

</LI>


<LI>
  [ftp, client] - Error when parsing a multiple FTP
response line. The last line in a multiple response must
be the response code followed by a space. A server may
have intermidate lines that start with the response code
even if this is not recommended. The parsing missed to
make sure that that space was present in what it
considered to be the last line.
<BR>
 Own Id: OTP-5712
<BR>

</LI>


<LI>
  [http, client] - The HTTP client will now retry a
pipelined request that was unsuccessful due to the fact
that the server unexpectedly closed the pipeline
connection.
<BR>
 Own Id: OTP-5728
<BR>

</LI>


<LI>
  [http, server, esi] - Under some circumstances mod_esi
would send a corrupted content-length header.
<BR>
 Own Id: OTP-5735
<BR>

</LI>


<LI>
  [http, server, get] - Removed debug printout which 
caused a confusing &#34;Socket closed&#34;-printout at high
load.
<BR>
 Own Id: OTP-5762 Aux Id: seq10101 
<BR>

</LI>


<LI>
  FTP: a data connection setup to the ftp server that
failed caused a crash of the client. Now it is handled
smothely.
<BR>
 Own Id: OTP-5738
<BR>

</LI>


<LI>
  [ftp] If host name is a ipv4 tuple ftp erronous tries to
connect as a ipv6 address with the ipv4 address.
<BR>
 Own Id: OTP-5764
<BR>

</LI>


<LI>
  [ftp] Handles connect to ipv6/ipv4 address differently
according to a change in <CODE>inet:getaddr</CODE>.
<BR>
 Own Id: OTP-5765
<BR>

</LI>


</UL>
<A NAME="1.10.2"><!-- Empty --></A>
<H4>1.10.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  [http, server] - The HTTP request handling was remoduled
to have a more straight forward error-handling. And the
internal debug strategy was changed to use tracing
instead of debug macros, which means we do not have to
write special debug code.
<BR>
 Own Id: OTP-5729
<BR>

</LI>


<LI>
  ftp:ls towards different ftp servers resulted in
different return results. E.g. the solaris 9 default
server caused <CODE>{ok,[]}</CODE> while older servers caused
<CODE>{error,epath}</CODE> as the result. For backwards
compatibility the behaviour has been changed to the old
result.
<BR>
 Own Id: OTP-5731
<BR>

</LI>


<LI>
  [http, server] - The documentation for the HTTP server
has been partly rewritten and very restructured too
provide a better overall picture. Lots of information
provided by &#34;semi manual pages&#34; has been moved to the
Users Guide.
<BR>
 Own Id: OTP-5752
<BR>

</LI>


<LI>
  FTP: verbose mode now also prints what the client sends
on the control channel.
<BR>
 Own Id: OTP-5753
<BR>

</LI>


</UL>
<A NAME="1.11"><!-- Empty --></A>
<H3>1.11 Inets 4.5.4</H3>
<A NAME="1.11.1"><!-- Empty --></A>
<H4>1.11.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [ftp, client] - Timing related issues could sometimes
cause the ftp response to be delayed.
<BR>
 Own Id: OTP-5705 Aux Id: seq10055 
<BR>

</LI>


<LI>
  [http, server, esi] - The dispatching of the post body to
the esi callback fuction was broken.
<BR>
 Own Id: OTP-5706
<BR>

</LI>


</UL>
<A NAME="1.12"><!-- Empty --></A>
<H3>1.12 Inets 4.5.3</H3>
<A NAME="1.12.1"><!-- Empty --></A>
<H4>1.12.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [ftp, client] - The FTP error code 550 was handled in an
unexpected way. Some earlier versions of inets had a
workaround for this in ftp:recv_bin/2 that was eliminated
during restructuring of the ftp module while implementing
ipv6 capabilities. The problem is now fixed.
<BR>
 Own Id: OTP-5682 Aux Id: seq10048 
<BR>

</LI>


<LI>
  [http, client] Post request with a body in binary format
failed as length was used instead of size.
<BR>
 Own Id: OTP-5686
<BR>

</LI>


<LI>
  [ftp, client] - For some FTP commands the FTP server will
send more than one reply on the FTP control channel. In
the case of a fast FTP server the client would sometimes
wrongly disregard the second answer as trailing garbage
attached to the first reply.
<BR>
 Own Id: OTP-5690 Aux Id: seq10055 
<BR>

</LI>


</UL>
<A NAME="1.12.2"><!-- Empty --></A>
<H4>1.12.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  [ftp, client] - A new option {progress, {CBmodule,
CBFunction, InitProgressTerm} has been added to allow
users to create things such as progress bars in there
GUI's. The option affects ftp:send/[3,4] and
ftp:recv/[3,4].
<BR>
 Own Id: OTP-5680
<BR>

</LI>


<LI>
  [http, client] - Added new API function http:request/1
<BR>
 Own Id: OTP-5691
<BR>

</LI>


<LI>
  [httpd, server] - mod_cgi is implemented according to
CGI-1.1 RFC 3875, an early implementation was based on
some draft that is not totally compliant to the RFC.
Documentation was updated. Also some code was
restructured to facilitate testing and maintenance of the
server.
<BR>
 Own Id: OTP-5694
<BR>

</LI>


</UL>
<A NAME="1.13"><!-- Empty --></A>
<H3>1.13 Inets 4.5.2</H3>
<A NAME="1.13.1"><!-- Empty --></A>
<H4>1.13.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [ftp, client] Calling ftp:recv/2 twice on the same
connection failed due to that the last message on the
ctrl channel was not appropriately taken care of. This
could potentially cause a problem for any operation
performed on the same connection where there had
previously been an ftp:recv/2 call. Also, in some cases,
when the process tries to close the data connection, it
does not take into account that the data connection may
actually not have been established.
<BR>
 Own Id: OTP-5662 Aux Id: seq10004, seq9988 
<BR>

</LI>


<LI>
  [ftp, client] Enhanced error handling, mainly so the ftp
client behaves gracefully when the user does strange
things such as violate the user API.
<BR>
 Own Id: OTP-5665
<BR>

</LI>


</UL>
<A NAME="1.13.2"><!-- Empty --></A>
<H4>1.13.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  [ftp, client] Added open option: mode.Deprecates
function force_active/1.
<BR>
 Own Id: OTP-5663
<BR>

</LI>


</UL>
<A NAME="1.14"><!-- Empty --></A>
<H3>1.14 Inets 4.5.1</H3>
<A NAME="1.14.1"><!-- Empty --></A>
<H4>1.14.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  [http, server] The server did not handle the config 
         directive BindAddress value &#34;*&#34; properly. 
         When creating the option list for the listen call, 
         everything beside the atom undefined (if BindAddress 
         was never given) and an 4-tuple (e.g. BindAddress
         value is 192.168.0.30) was incorrectly assumed to be 
         an ipv6 address. 
         For undefined (no BindAddress), Inets attempts to 
         figure out if it is running on a ipv6-machine, and if
         so, add the inet6 option when calling listen. The 
         same approach should be used when BindAddress is 
         assigned the value &#34;*&#34;.
<BR>
 Own Id: OTP-5642
<BR>

</LI>


<LI>
  Some data doesn't pass through http_base_64:decode/1 correctly.
The decoding routine fails whenever a 4-character group of the 
encoding ends with &#34;9&#34; or &#34;99&#34;. If it ends with 99, two bytes 
will be lost in the decode routine. If it ends with a single 9, 
one byte will be lost in the decode.
<BR>
 Own Id: OTP-5635 Aux Id: seq9971
<BR>

</LI>


<LI>
  [http, server,esi] Web server does not handle status-code 
returned by an esi function. I.e. the esi-function 
can no longer control the status code.
<BR>
 Own Id: OTP-5648 Aux Id: seq9982
<BR>

</LI>


<LI>
  [http, server,esi] Corrected header format. First character 
was lower case, and there where no space after the &#34;:&#34;
character, example: content-length:0. Now, first character 
was upper case, and a space after the &#34;:&#34; character, 
example: Content-Length: 0 (To preserve
backward compatibility with the de-facto standard as the
new way does not break the HTTP standard!)
<BR>
 Own Id: OTP-5649 Aux Id: seq9982
<BR>

</LI>


<LI>
  [http, server, cgi] Parsing of the status header field could
cause a crash.
<BR>
 Own Id: OTP-5650 Aux Id: seq9982
<BR>

</LI>


</UL>
<A NAME="1.15"><!-- Empty --></A>
<H3>1.15 Inets 4.5</H3>
<A NAME="1.15.1"><!-- Empty --></A>
<H4>1.15.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  The internal design of using blocking gen_tcp:recv with a
timeout and retries resulted in code that was hard to get
a good overview of, and ultimate led to situations where
the client got the wrong answer or no answer at all. The
errors where many times very timing dependent and mainly
effected the chunk-related functions, so if you where
lucky you proably would not have noticed. The internal
design was changed to use gen_tcp active once semantics.
The API is not effected except for the function
ftp:quote/2 which now returns a list of strings (ftp
result lines) where the line endings &#34;\r\n&#34; has been
removed. This was the original intention for the return
value of ftp:quote/2 but it was non trivial to make a
good such solution with the old design and a compromise
was made.
<BR>
 *** POTENTIAL INCOMPATIBILITY ***
<BR>
 Own Id: OTP-5623
<BR>

</LI>


<LI>
  The new implementation of pipelining in inets-4.1 alas
was slightly broken and unfortunately not caught by the
test suite that apparently needs some additions. The
result was that requests that ought to have been
pipelined where not, this has now been fixed.
<BR>
 Own Id: OTP-5624
<BR>

</LI>


<LI>
  When using the latest esi interface with the callback
interface of arity 3, HTTP Content-Type headers where
ignored, this due to a subtle difference between this
interface and the old one in how they viewed HTTP
delimiters.
<BR>
 Own Id: OTP-5627
<BR>

</LI>


</UL>
<A NAME="1.15.2"><!-- Empty --></A>
<H4>1.15.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  The HTTP server now supports ipv6 in the case that the
underlying mechanisms also do so. (ssl does not yet
support ipv6.)
<BR>
 Own Id: OTP-5141
<BR>

</LI>


<LI>
  The FTP client now supports ipv6 in the case that the
underlying mechanisms also do so.
<BR>
 Own Id: OTP-5142
<BR>

</LI>


<LI>
  An option was added to disable the ipv6 support in the
HTTP client. This to provide a workaround possibility for
buggy ipv6-stacks.
<BR>
 Own Id: OTP-5625 Aux Id: seq9872
<BR>

</LI>


<LI>
  When generating dynamic HTTP response bodies the the
default content-type is now set to &#34;text/html&#34; instead of
&#34;text/plain&#34; which is more intuitive.
<BR>
 *** POTENTIAL INCOMPATIBILITY ***
<BR>
 Own Id: OTP-5626
<BR>

</LI>


</UL>
<A NAME="1.16"><!-- Empty --></A>
<H3>1.16 Inets 4.4.1</H3>
<A NAME="1.16.1"><!-- Empty --></A>
<H4>1.16.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  Wrapper function http_transport:accept/3 did
error-handling that it should not do because only the
caller of the wrapper can determine what action to take.
So timeouts where handled twice, once in http_transport
and once in httpd_acceptor. Clean up of the wrapper
module http_transport changed the action of the wrapper
module and made the unwanted behavior noticeable in in
OTP error logs. And now the unwanted error handling has
been removed. The cleanup helped us find bad code but
alas it also generates a lot of log printouts that are
quite disturbing to the user of the HTTP-server.
<BR>
 Own Id: OTP-5549 Aux Id: seq9851 
<BR>

</LI>


<LI>
  In the rewrite for 4.4 some mod_esi-environment values
where mistaken for ordinary header values and where
incorrectly transformed to strings. They are now atoms
again.
<BR>
 Own Id: OTP-5551 Aux Id: seq9854 
<BR>

</LI>


<LI>
  The HTTP server now handles &#34;GET /\r\n\r\n&#34; as well as
&#34;GET / \r\n\r\n&#34;. According to the RFC the whitespace is
not nedded.
<BR>
 Own Id: OTP-5552 Aux Id: seq8426 
<BR>

</LI>


</UL>
<A NAME="1.16.2"><!-- Empty --></A>
<H4>1.16.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  The ftp client now supports passive mode. Actually the
ftp client will always try to use passive mode and if it
fails it will use active mode instead. It is also
possible to force the ftp-client to use active mode, if
that is desired, by calling ftp:force_active/1 this way
you can get the old behavior.
<BR>
 *** POTENTIAL INCOMPATIBILITY ***
<BR>
 Own Id: OTP-5148
<BR>

</LI>


</UL>
<A NAME="1.17"><!-- Empty --></A>
<H3>1.17 Inets 4.4</H3>
<A NAME="1.17.1"><!-- Empty --></A>
<H4>1.17.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  The server did not handle HTTP-0.9 messages with an
implicit version.
<BR>
 Own Id: OTP-5513
<BR>

</LI>


<LI>
  An internal server timeout killed the request handling
process without sending a message back to the client. As
this timeout only affects a single request it has been
set to infinity (if the main server process dies the
request handling process will also die and the client
will receive an error). This might make a client that
does not use a timeout hang for a longer period of time,
but that is an expected behavior!
<BR>
 Own Id: OTP-5514 Aux Id: seq9806 
<BR>

</LI>


<LI>
  That a third party closes the http servers accept socket
is recoverable for Inets, hence intes will only produce
an info report as there was no error in Inets but
measures where taken to avoid failure due to errors
elsewhere.
<BR>
 Own Id: OTP-5516 Aux Id: seq9806 
<BR>

</LI>


<LI>
  The HTTP client proxy settings where ignored. Bug
introduced in inets-4.3.
<BR>
 Own Id: OTP-5517
<BR>

</LI>


<LI>
  Inets only sent the &#34;WWW-Authenticate&#34; header at the
first attempt to get a page, if the user supplied the
wrong user/password combination the header was not sent
again. This forces the user to kill the browser entirely
after a failed login attempt, before the user may try to
login again. Inets now always send the authentication
header.
<BR>
 Own Id: OTP-5521
<BR>

</LI>


<LI>
  A major rewrite of big parts of the HTTP server code was
performed. There where many things that did not work
satisfactory. Cgi script handling can never have worked
properly and the cases when it did sort of work, a big
unnecessary delay was enforced. Headers where not always
treated as expected and HTTP version handling did not
work, all responses where sent as version HTTP/1.1 no
matter what.
<BR>
 Own Id: OTP-5537
<BR>

</LI>


</UL>
<A NAME="1.18"><!-- Empty --></A>
<H3>1.18 Inets 4.3.1</H3>
<A NAME="1.18.1"><!-- Empty --></A>
<H4>1.18.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  When further testing the functionality of https requests
that goes through a proxy. We realised that alas this can
not currently be supported as it requires features from
the ssl implementation that is not currently available.
So for now an error message will be returned when trying
to use this functionality.
<BR>
 Own Id: OTP-5453
<BR>

</LI>


<LI>
  When trying to get a url from a server that does not
exist the client hanged instead of returning an error
message. Bug introduced in inets-4.3.
<BR>
 Own Id: OTP-5454
<BR>

</LI>


</UL>
<A NAME="1.19"><!-- Empty --></A>
<H3>1.19 Inets 4.3</H3>
<A NAME="1.19.1"><!-- Empty --></A>
<H4>1.19.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  Tunneling of SSL through a proxy has now been
implemented. However due to lack of test sites this has only
partially been verified, it is likely that there will
have to be future improvments in this area.
<BR>
 Own Id: OTP-5443
<BR>

</LI>


</UL>
<A NAME="1.19.2"><!-- Empty --></A>
<H4>1.19.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  The pipeline timeout was changed to be zero by default to
avoid that people by accident would create connection
processes that never dies and eats up the socket
resources.
<BR>
 *** POTENTIAL INCOMPATIBILITY ***
<BR>
 Own Id: OTP-5442
<BR>

</LI>


<LI>
  Altered the way spawn_link is used in mod_esi to avoid
getting, in this senario unwanted error reports, from
spawn_link. (The behavior of spawn_link was altered in a
not backwards compatible way.)
<BR>
 Own Id: OTP-5444
<BR>

</LI>


</UL>
<A NAME="1.20"><!-- Empty --></A>
<H3>1.20 Inets 4.2.1</H3>
<A NAME="1.20.1"><!-- Empty --></A>
<H4>1.20.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  Sometimes EWS modules where called with an Info record
where the peername field was {-1, &#34;unknown&#34;}. This could
happen when a client was making a lot of requests which
it discards before they where answered by the server. The
server now ignores such requests and does not call the
EWS modules in this case.
<BR>
 Own Id: OTP-5380 Aux Id: seq9739 
<BR>

</LI>


<LI>
  The HTTP-server now returns the 408 status code upon a
request timeout as expected instead of the previous
faulty behavior of sending a 500 status code.
<BR>
 Own Id: OTP-5409
<BR>

</LI>


<LI>
  The content length was put in to the HTTP-headers as an
integer instead of as a string.
<BR>
 Own Id: OTP-5410
<BR>

</LI>


<LI>
  It was wrongly presumed that code:priv_dir would always
be writable due to how the test-server works. The
directory is now a configuration parameter in the
inets-application configuration file. Failing to
configure it will result in that all cookies are treated
as session cookies.
<BR>
 Own Id: OTP-5411
<BR>

</LI>


</UL>
<A NAME="1.20.2"><!-- Empty --></A>
<H4>1.20.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  An undocumented beta version of tftp is included.
<BR>
 Own Id: OTP-5419
<BR>

</LI>


</UL>
<A NAME="1.21"><!-- Empty --></A>
<H3>1.21 Inets 4.2</H3>
<A NAME="1.21.1"><!-- Empty --></A>
<H4>1.21.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  When sending a request through a proxy the absolute URI
must be used.
<BR>
 Own Id: OTP-5368
<BR>

</LI>


</UL>
<A NAME="1.21.2"><!-- Empty --></A>
<H4>1.21.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  Basic support for cookies was implemented. Later some
more functions to inspect cookies may be added.
<BR>
 Own Id: OTP-5331
<BR>

</LI>


<LI>
  A top tftp supervisor was added in preparation for adding
a tftp service in a future Inets release.
<BR>
 Own Id: OTP-5379
<BR>

</LI>


</UL>
<A NAME="1.22"><!-- Empty --></A>
<H3>1.22 Inets 4.1</H3>
<A NAME="1.22.1"><!-- Empty --></A>
<H4>1.22.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  The URI check that disables relative links that goes
outside the server-root still missed a few cases, in
spite of the improvement in OTP-5140.
<BR>
 Own Id: OTP-5249
<BR>

</LI>


<LI>
  The http client pipelining implementation has been
rewritten as the old implementation was too optimistic
about when to pipeline. In the process of doing this also
the error handling was improved, better clean up is
performed when the request handling process terminates
and better handling of the case that the httpc_manager
process dies and is restarted.
<BR>
 Own Id: OTP-5303
<BR>

</LI>


<LI>
  Improved handling of status codes 30X and 50X.
<BR>
 Own Id: OTP-5309
<BR>

</LI>


</UL>
<A NAME="1.22.2"><!-- Empty --></A>
<H4>1.22.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  The Inets supervision tree has been reorganized to create
a better balance between the Inets services. Preferably
they should not effect each other. The ftp service has
also been included in the Inets supervision tree, it was
for reasons unknown, not included before.
<BR>
 Own Id: OTP-5188
<BR>

</LI>


<LI>
  The service concept in Inets is now better documented.
<BR>
 Own Id: OTP-5189
<BR>

</LI>


<LI>
  The Inets shutdown times have proven to be too short
under some circumstances, as a heavy load, therefore they
have been prolonged.
<BR>
 Own Id: OTP-5261 Aux Id: seq9624
<BR>

</LI>


<LI>
  Options for automatic redirection and pipelining is now
available in the http client API.
<BR>
 Own Id: OTP-5304
<BR>

</LI>


</UL>
<A NAME="1.23"><!-- Empty --></A>
<H3>1.23 Inets 4.0.1</H3>
<A NAME="1.23.1"><!-- Empty --></A>
<H4>1.23.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  A programming error could cause a badmatch in the
http-client when the http response was chunk decoded.
<BR>
 Own Id: OTP-5101
<BR>

</LI>


<LI>
  The parsing of HTTP messages was missing a base case.
This caused unexpected behavior when the separator CR and
LF where received in different tcp packets.
<BR>
 Own Id: OTP-5239
<BR>

</LI>


</UL>
<A NAME="1.24"><!-- Empty --></A>
<H3>1.24 Inets 4.0</H3>
<A NAME="1.24.1"><!-- Empty --></A>
<H4>1.24.1 Fixed Bugs and Malfunctions</H4>

<P>
<UL>

<LI>
  When receiving a status 100 code, the client should only
respond by sending the message body, if the client sent
an expect header in the first place. Failing to do so may
result in that the server receives the body twice.
<BR>
 Own Id: OTP-4848
<BR>

</LI>


<LI>
  mod_get now also handles http version HTTP/0.9
<BR>
 Own Id: OTP-4935 Aux Id: seq8426 
<BR>

</LI>


<LI>
  &#34;Last-modified&#34; field was incorrectly set to local time
with the tag GMT, it is now corrected so that the time
reflected is in fact GMT.
<BR>
 Own Id: OTP-4936
<BR>

</LI>


<LI>
  The client will only add a host-field to the request if
there is not one already present.
<BR>
 Own Id: OTP-4984
<BR>

</LI>


<LI>
  The Inets application tries to be compatible with
Apache. To be more compatible the option
'MaxKeepAliveRequest' is renamed 'MaxKeepAliveRequests'.
The old name is kept for backward compatibility.
<BR>
 Own Id: OTP-5024
<BR>

</LI>


<LI>
  Changing the base 64 decoding to not accept invalid
input, uncovered a logical error in mod_security.erl An
already decoded string was sent as input to decode. In
this case, as it so happened, the two errors worked
together creating the elution that everthing was right.
This has now been corrected.
<BR>
 Own Id: OTP-5083
<BR>

</LI>


<LI>
  URLs where not properly scrutinised for relative paths. A
malicious user could exploit this to read files outside
the document root. This is no longer the case.
<BR>
 Own Id: OTP-5140
<BR>

</LI>


</UL>
<A NAME="1.24.2"><!-- Empty --></A>
<H4>1.24.2 Improvements and New Features</H4>

<P>
<UL>

<LI>
  A HTTP 1.1 client is officially included in Inets. It is
loosely based on the previously unsupported code
contributed by Johan Blom. In this first version only the
most basic HTTP functionality is supported. The user API
has been changed.
<BR>
 *** POTENTIAL INCOMPATIBILITY ***
<BR>
 Own Id: OTP-5047
<BR>

</LI>


<LI>
  Fixed erroneous link in documentation.
<BR>
 Own Id: OTP-5089 Aux Id: seq8887 
<BR>

</LI>


<LI>
  Added the function quote/2 that lets you send an
arbitrary FTP command to the FTP client.
<BR>
 Own Id: OTP-5099 Aux Id: seq8961 
<BR>

</LI>


<LI>
  Started integration of the HTTP client and server code
too facilitate maintenance and further development.
<BR>
 Own Id: OTP-5110
<BR>

</LI>


<LI>
  Due to several possibilities to interpret the ftp
standard some newer ftp-servers have interpreted the
standard in such a way that the documented return value
of ftp:nlist/2 does not always match the actual return
value. Some extra checks have now been added to ensure
the documented return value. This will also result in
that ftp:nlist is not bug compatible in the case that
nlist is given a filename instead of a directory it will
now return an error instead of {ok, FileName}.
<BR>
 *** POTENTIAL INCOMPATIBILITY ***
<BR>
 Own Id: OTP-5165
<BR>

</LI>


<LI>
  Created a Users Guide for Inets. Earlier there where some
fake manual pages and information was scattered
everywhere and hard to find.
<BR>
 Own Id: OTP-5180
<BR>

</LI>


</UL>

<P> For information about older versions see
<A TARGET="_top" HREF="part_notes_history_frame.html">release notes history</A>.
<CENTER>
<HR>
<SMALL>
Copyright &copy; 1991-2006
<A HREF="http://www.erlang.se">Ericsson AB</A><BR>
</SMALL>
</CENTER>
</BODY>
</HTML>