*******************************************************************************
****This is README for tmview v98.06  (C)opyright 1995 Thomas Moor*************
*******************************************************************************

What is tmview ?  tmview is a screen-previewer for .dvi-files compiled by
TeX. It let's you see what your printed output will look like. You can choose
between a black-and-white representation and greyscaling. You can choose an
arbitrary zoomfactor (at some cost of performance). You can set marks to
measure distances. You can search for textstrings. You may visit lots of
DVIfiles, set bookmarks and get them saved to a startup-file. tmview does not
support pxl-files since I think they are prehistoric. tmview ignores all 
'special'-commands, sorry lads. tmview tries its best with virtual fonts and
is prepeared to be linked with kpathsea.

The current version goes with linux/svgalib as well as
the X Window System.


*******************************************************************************
USAGE**************************************************************************
*******************************************************************************

tmview [-?] [-h<marg>] [-v<marg>] [-p<width>x<height>] [-r<xres>x<yres>] 
       [-f<path>] [-n<name>] [-t<path>] [-g<path>] [-d<width>x<height>]
       [-m<mag>] [-k<leftmarg>,<rightmarg>,<uppermarg>,<lowermarg>]
       [-s<startup>] [<file>[.[dvi]]]

If you call tmview with the option -?, tmview will give a complete list of 
all possible options and their default-values, after reading the startup-
file, if any.

IMPORTANT: All options have to be followed IMMEDIATELY by their arguments,
NO spaces must be inserted !!!



COMMANDLINE-OPTIONS************************************************************

The command-line-options are meant to prepare tmview for going with your
systems TeX-setup. While the defaults work fine on my system, you may have to
pass suitable values. See -s for keeping your defaults in a startup-file.

-?     (Help command-line-options)
       List all command-line-options and their defaults, after reading the 
       startup-file, if any. (see -s).

-h     (Horizontal-offset)
       A lot of printer drivers do a horizontal offset of 1 inch. If yours
       does as well, you should use "-h25.4". The length following -h must be 
       given in mm.
       Default: "25.4"                                             

-v     (Vertical-offset)
       A lot of printer drivers do a vertical offset of 1 inch. If yours
       does as well, you should use "-v25.4". The length following -v must be 
       given in mm.
       Default: "25.4"                                              

-p     (Paper-size)
       Tells tmview width and height of the paper you are using. The width is 
       given first and both width and height are given in mm. Width and height
       are separated by an "x".
       Default: "210.0x297.0" (German DIN A4)                 

-r     (Resolution)
       Tells tmview what kind of pk-files to use. The horizontal resolution is 
       given first and both horizontal and vertical resolution are given in
       dpi (dots per inch). The two values are separated by an "x". Saying 
       "-r600x600" means tmview will use fonts which were generated for a 
       600-dpi-printer. As tmview assumes that pixels on the screen are square,
       different values for horizontal and vertical resolution will result in 
       a distorted image. Always make sure the desired pk-files are available.
       The chosen resolution determines the (maximum) size of the 
       representation (comp. <+> below).
       Default: "300x300"                                     

-f     (Font-path)
       A list of paths telling tmview where to look for the pk-files. The items
       in this list have to be separated by ":"'s. The given list is executed
       from left to right. If an item ends with "//" all subdirectories will be
       scanned too. This is programmed in a very odd way, so it takes lots of
       time. You should place such items at the end of the list.
       Default: "./:/usr/lib/texmf/fonts//"                   

-n     (Name-of-the-font-file)
       If your DVIfile tells tmview to use a font called "thisnthatfont" and 
       tmview was told to use a resolution of "123" dpi, tmview has to know 
       how the desired file is named. 
       In the string following "-n" the following replacements are made:
       xxx     replaced by

       @N      thisnthatfont
       @K      thisntha  (this is @R reduced to 8 characters (MS-DOG !))
       @M      123       (the resolution)
       @R      615       (this is @M*5, intended for magnified 200 dpi 
                          fonts used instead of 300 dpi fonts)  
       Defaults: "@N.@Mpk"  

-t     (Tfm-path)
       A list of paths telling tmview where to look for the tfm-files. The 
       items in this list have to be separated by ";"'s. The given list is 
       executed from left to right. tfm-files are used, when tmview can't
       find a fonts pk-file. In that case boxes are drawn instead of the 
       glyphs. 
       Defaults: "./:usr/lib/texmf/fonts//"  
 
-g     (Vf-path)
       A list of paths telling tmview where to look for the vf-files.
       The support of vf-files is not very well tested. If anything
       goes wrong with vf-files, use dvicopy to replace them by
       ordinary dvi-code. Nomenclature of the list is as above.
       Default: "./:usr/lib/texmf/fonts//"
 
-d     (Display)
       Tells tmview the desired size of your display. The width is given 
       first and both width and height are given in pixels. Width and height 
       are separated by an "x". Using linux you may choose a resolution
       supported by your version of svgalib with respect to your hardware. 
       Only 256-color-modes are allowed. That means to get more than 320x240 
       you must have a supported svga chip-set, standard vga wont do. 
       See INSTALLATION on how to run tmview with standard vga. 
       Default: look for the value of $GSVGAMODE if set, 
                otherwise try "640x480".       
       
-m     (Magnification)
       If you want to magnify by a factor n you have to specify n*1000 as an 
       argument to "-m", e.g. "-m2000" means all lengths will be doubled. 
       NOTE: tmview magnifies according to the origin of the dvi-coordinates,
       which is -in most cases- NOT the upper left corner of the paper. Always
       make sure the desired pk-files are available. "-m" doesn't magnify the
       paper, so if you wish a larger image use the "+" and "-" keys or the
       "-r" option.
       Default: get magnification from DVIfile.

-k     (Kannot-print-any-further)
       Most printers stop printing if they are too close at the papers edge 
       (some might even do strange things). The -k option describes the 
       printable area, e.g. saying "-k1.0,2.0,3.0,4.0" means that your printer
       can print as close as
       1 mm to the left
       2 mm to the right 
       3 mm to the upper 
       4 mm to the lower
       edge of the paper. These values are used to draw a frame indicating the
       printable area (comp. <p> below). When searching textstrings, any text
       outside the printable area is ignored. All four values have to be given
       and they have to be separated by ",".  All lengths are given in mm.    
       Default: "4.0,4.0,4.0,12.0"      

-s     (startup-file) 
       The above options, a list of visited DVIfiles (see bookmarks below) and
       lots of other interna may be loaded from the startup-file, specified
       directly after the -s.  When quitting tmview the current options
       etc. will be saved in the startup-file. If this is not desired, the
       startup-file has to be set read-only !!  A possible tactic therefore it
       is, to run tmview with suitable options, vistit often used DVIfiles, set
       within theese desired manual-bookmarks and save all that information in
       the startup-file. Then set it read-only and/or copy it somewhere
       save. See also editing the startup-file below.  
       Default: "~/.dvisvga"



*******************************************************************************
VIEWING A DVI-FILE*************************************************************
*******************************************************************************

When everything is setup right, visiting a DVIfile with tmview just means to
navigate the visible area along that file, using the cursor-keys.  Some of the
following commands obviously require an argument, f.e. <g>. Arguments are to
be entered BEFORE executing a command. When executed by hitting the
corresponding uppercase key, you will be asked for the argument.  While the
user-interface of tmview is meant to save keystrokes, it is not too
intuitive. You may either read the following instructions, or just do
<caps-lock>.


<?>               give a short description of the following command,
                  hitting <?> twice gives the list of all commands 

<i>/<m>           select a page relative to the current page 
<g>               select a page w.r.t. TeX-counters
<u>/<n>/<h>/<j>   scroll the visible area
<f>/<c>           make scrolling finer/coarser
<z>               center visible area
<+>/<->           zoom in/out
<v>               set zoom-factor

<b>		  set a bookmark
<w>               move to a bookmark
<^>               move back

<o>               toggle greyscaling
<x>               toggle statusline-information
<t>               set unit of measurement

<l>               show/hide screenmark and pagemark
<y>               set pagemark at the position of screenmark
<a>               show/hide marked rectangle
<p>               show/hide printable area
<e>               set page-offset and -size

<k>               show/hide half-hyper-tex-mark
<TAB>             move to next href
<RET>             follow current href

<s>               search for text
<r>               re-read DVIfile and re-draw screen
<d>               load/kill DVIfile
<q>               quit tmview


The cursor-keys <page-up> and <page-down> are taken as <i> resp. <m>. The
keys <pos1> and <end> select the first resp. the last page. The cursor-keys
<left> <right> <up> and <down> do <h> <j> <u> resp. <n>. All in all this means,
that the cursor-keys do what they are meant to.

When a mouse is connected, it moves any visible mark. When the right or left
mousebutton is no good for anything else, moving the mouse while holding that
button acts on the visible area instead. (That sounds bad, but works out fine)
When the screenmark is shown (see <l>), the left mouse-button sets the
pagemark (see <y>) and the right button marks a rectangle (see <a>). When the
half-hyper-mark is shown, the left mouse-button follows the current href, if
any.



******************************************************************************
NON-INTUITIVE USER-INTERFACE **************************************************
******************************************************************************

To explain the way tmview expects to receive commands, a short nomenclature
used in the sequel is given:
 
Any text within `<' and `>' represents a single keystroke, while `(' and `)'
mark the beginning resp. the end of a string to enter.  So <h> is about to hit
the key named `h' and (25.4) could stand for <2><5> <.><4>.  The
carriage-return-key is referred by <ret>, the backspace-key or delete-key by
<del> and the escape-key by <esc>. Almost whenever a string is expected by
tmview, you may use <del> to remove the last character you typed. So even
<2><5><6><del><.><4> results in (25.4). Note that a string does not need to be
terminated by <ret>.  To simplify a reference to a string, in this text any
uppercase letters within `(' and `)' are not meant as a string, but as the
name of some string. So we may read something like `(PP)<h> scrolls PP percent
to the left' as `typing in any number just before <h> results in scrolling left
according to that number'.



COMMANDS AND ARGUMENTS*********************************************************

A command is executed by typing its name, which consists of a single
character. Some of the above listed commands take numerical arguments.
Arguments are always optional. They may be entered before executing the
command. Multiple arguments are separated by <,> or <;>. If no argument is
passed, a default is used. If an argument is passed, it serves as the default
for following commands. Commands doing similar things share the same default
arguments.

Example: 
Typing (10)<h> results in scrolling 10% to the left and sets the default for
any scrolling commands to 10%. Thus typing <j> afterwards results in scrolling
10% to the right.



MAGIC ARGUMENT*****************************************************************

As a special argument some commands accept the magic argument <*>. It is used
either to vary the command in some way or to get the arguments from another
place. <#> is equivalent to <*> and saves you from holding down the shift-key
on some keyboards.

<z> for example centers the visible area. It excepts as argument the point
which will be taken as origin. Since this will become the default for a future
<z>, you may measure out the origin only once. Even quicker it is to position
the visible area by scrolling and then to do a <*><z>.  This results in taking
the current position as centered and so as default for a coming <z>.

As another example look at the command <m>, which moves pages forward,
according to a given argument resp. a default. By moving on the next page
there are two modi available: 1. keep the visible area; 2. do center like
<z><z>.  <*><m> toggles between these two modi. So in this case <*> acts as a
kind of flag argument.



SELECTING THE PAGE************************************************************

<i>/<m> Select a page relative to the current page. 

     <i> moves towards the beginning of the DVIfile, while <m> moves towards
     the end. A single argument (N) may specify the amount of movement in
     pages. If the argument is (*), the page-moving-mode is toggled. See
     above. 
     Example:
     (1)<i> selects the previous page 
     <m>    selects the next page

<g> Select a page with respect to TeX counters.
   
     A list of ten arguments (COUNT0; COUNT1; ... COUNT9) specifies the page
     to be selected. <*> may be used as wildcard. If there are more than one
     but less than ten arguments given, the others will be taken as <*>. 
     Example:
     (26)(i) selects the first page after the current page with a value of 
             26 in \count0.  



MOVING AROUND ON CURRENT PAGE *************************************************

<u>/<n>/<h>/<j> Scrolling the visible area

       A single argument (PP) may be used to specify the amount of scrolling
       in percent of the screen-width. <u>/<n> resp. <up>/down> both accept the
       argument (*) to toggle between: 1. stay on the current page; 2. scroll
       over pages. When scrolling over pages, you may view the whole document
       while using only the single key <n>.
       Example:
       (20)<h> scrolls 1/5 to the left.  

<f>/<c> Make scrolling finer/coarser

       These commands change the default argument for the above scrolling-
       commands. So <f> and <c> don't move the visible area at all, but they
       change the way the scrolling-commands act. 
    
<z> Center the visible area

       Without any argument <z> centers horizontally only, while <z><z> 
       centers in both directions. 
       When two arguments (X,Y) are given, they describe the point on the
       page, which will become the middle of the visible area. When the
       argument (*) is given, the current position is taken as centered.
       When the screenmark is shown (see below <l>), and the argument (*) is
       given, the position of the screenmark becomes the center. This does
       move the visible area, but it does not move the screenmark. 
    



ZOOMING *******************************************************************


<v> Set the zoom-factor
 
       This command requires one argument (F) which must be between 0.1 and 2.
       F will become the zoom-factor and the visible-area will be redrawn. If
       the zoom-factor is 1, the pixels found in pk-files are just copied one
       by one to the screen. Since the resolution of our days screens seems to
       be less than that of our days printers, and since you may still want to
       use the same pk-files for printing and viewing, F=1 usually results in
       a magnification. So when you're just reading some text in some DVIfile
       you will set F to something like 0.3, depending on the involved
       resolutions. When the screenmark is visible (see <l>) the position of
       the screenmark is taken as the origin of zooming, i.e. it is fixed. When
       the screenmark is not visible, the middle of the screen is fixed. The
       current zoom-factor is displayed in the optional statusline, see <x>.

       There are two zooming modi. The fast modus works only if 1/F is an
       integer.  So good values are F=0.5, 0.333, 0.25, 0.2, 0.167 and so
       on. When in the fast modus, the <+>/<-> keys described below step
       through that values. And since F is to be small anyway, you may never
       use <v> to set it explicitly.

       The good thing about the slow modus is, that it allows you to choose
       the zoom-factor arbitrary (between 0.1 and 2). So poor students with
       small screens might find some optimum to make the text fit and still be
       readable. The bad thing about the slow modus is that it is slow.  But
       since once zoomed glyphs are kept in memory, this slowness only hurts
       the first few pages after changing the zoom-factor. Anyway, <*><v>
       brings you back in the fast modus, choosing the nearest possible
       zoom-factor therefore.


<+>/<-> Zoom in/out

       Increase/decrease the zoom-factor. When in the fast modus, step through
       the fast values only (see above). When an argument (PP) is given, it
       is taken as the amount of increasing/decreasing in percent of the
       current zoom-factor. This is likely to result in the slow modus.

       Example:
       (0.17)<v>   set zoomfactor to 0.17               (slow)
       (10)<+>     set zoomfactor to 1.1*0.17=0.187     (slow)
       (3)<->      set zoomfactor to 0.97*1.87=0.181    (slow)
       <->           ...          to 0.97*0.181=0.176   (slow)
       (*)<v>        ...          to 1/6=0.167          (fast)
       <+>           ...          to 1/5=0.2            (fast)
             

BOOKMARKS *********************************************************************

A bookmark remembers what is seen on the screen. That is the DVIfile, the page
within that file, the position of the visable area and the zoom-factor. There
are three kind of bookmarks ...

file-bookmarks:
Each file visited has a file-bookmark, containing the above information about
what was seen on the screen when visited the last time, plus some information
on the file, that is the paper-offset and -position, the location of the
printable-area.  file-bookmarks are generated automaticly.  This results in
easy re-visiting a DVIfile: you'll find it as left. A file-bookmark is removed
by killing the DVIfile with <d><k>, see <d> below.

back-bookmarks:
When searching a text-string, following a href or moving to a bookmark, the
position within the DVIfile might be changed to somewhere far far away. To
simplify recovering fromsuch excursions, a back-bookmark will be generated
automaticly.  To prevent getting fed up with thousands of back-bookmarks, the
total number of theese is limited. See <^> below.

manual-bookmarks:
After all you may install your own bookmarks, marking often visited places,
say in some manuals. manual-bookmarks are named by a number.  This number has
to be unique whithin the DVIfile they belong to.  To define a manual-bookmark
use <b>. Since manual-bookmarks belong to the DVIfile they are defined on,
they get lost, when that DVIfile is killed by <d><k>. 

All kind of bookmarks are kept in a ring-buffer. There is a so called current
bookmark of each type. Visiting the bookmarks along the ring-buffer is done by
<w> for file- and manual-bookmarks, while <^> acts on back-bookmarks.

<b> Define/undefine manual-bookmark.

       When the current position is not already defined as a manual-bookmark,
       <b> defines one. When an single numeric argument (NUM) is given, NUM
       will be the name of the newly defined bookmark. With no argument, a
       name will be generated automaticly. See <^> below, for how to visit
       manual-bookmarks.  When the current position is already defined as a
       manual-bookmark, <b> undefines that manual-bookmark.

<w> Move to bookmark.

       When a single numeric argument (NUM) is given, <w> moves to the
       manual-bookmark named NUM, if any. Since manual bookmarks are bound to
       DVIfiles, the current DVIfile will never change in that case. If no
       argument is given, <w> goes moves the postion either thrue the
       ring-buffer of file-bookmarks or thrue the one of manual-bookmarks. To
       toggle between theese two modi, use the magic argument <*>.

<^> Move back
   
       Move to the latest back-bookmark, if any. When a single numeric
       argument (TOTAL) is given, keep the TOTAL latest back-bookmarks and
       discard all the others.



CHOOSING WHAT'S ON THE DISPLAY ************************************************

<o> Toggle greyscaling
 
       When the zoom-factor is less than 1, the glyphs may be displayed using
       grey-levels, making them more smooth. This takes some memory, so you
       may switch it off. On high-res displays there is no need for it anyway.
 
<x> Toggle statusline-information

       While the standard statusline shows you the page-number of the current
       page and the arguments you are about to enter, you may select optional
       information for measuring out distances and so. See below.

<t> Set unit of measurement

       Whenever you specify arguments which are to describe a point on the
       page, this is done w.r.t. a unit of measurement, i.e. cm, mm,
       a.s.o.. This unit is also used, when the position of a mark is
       displayed in the statusline.



MEASURING *********************************************************************

To allow you to measure distances on the page, there are two marks, the 
screenmark, which is fixed on the physical screen you're looking at, and the
pagemark, which is fixed on the DVIfiles page. When you move the visible-area,
the screenmark acts as drawn with edding on your monitor. The pagemark acts as
drawn on the page. The optional statusline tells the position of the pagemark
relative to the corner of the sheet of paper you're viewing. It also tells
the position of the screenmark relative to the pagemark. To measure distances
you first may switch this marks on, using <l>. When the marks are shown,
the scrolling commands don't act on the visible area anymore, but move the 
screenmark. For that case only moving the screenmark at the boarder of the 
screen results in scrolling. To move the pagemark just move the screenmark at 
the desired position and use <y> to make the pagemark follow.
 
<l> Show/hide screenmark and pagemark

       This commands takes the two arguments (PM_X;PM_Y). The pagemark is put
       at position PM_X PM_Y w.r.t. the upper left corner of the page. The 
       Screenmark may be moved with the scrolling-commands.

<y> Set pagemark at the position of the screenmark
       
Beside of these marks there are three rectangles for measurement. First there
is the boarder of the paper setup by the command-line options -h,-v and -p.
Then there is the printable area, setup with the -k command-line option.
Third the so called marked rectangle used.

<a> Show/hide marked rectangle

       The four arguments (LEFT,TOP,WIDTH,HIGHT) specify the position on page
       an the size of the marked rectangle. When pagemark and screenmark are
       shown, their positions are used as default. When they are hidden, the
       last position of the marked rectangle is used as default.
 
<p> Show/hide printable area 
       
       The four arguments (LEFT,RIGHT,TOP,BOTTOM) specify the margins
       of the printable area, w.r.t. the boarder of the page. When pagemark
       and screenmark are shown, the argument (*) sets the printable area to
       the rectangle described by screenmark and pagemark.  When they are
       hidden, (*) takes the command-line-options resp. defaults -k of the
       printable area.

<e> Set paper-offset and -size  

       The four arguments (HOFF;VOFF;WIDTH;HEIGHT) specify the boarder of the
       page. Have the top-left corner of a sheet of paper in mind. Then
       (HOFF,VOFF) is the offset of the DVIfile's origin to the left boarder
       of the paper.  Standard values are HOFF=VOFF=2.54cm. WIDTH and HEIGHT
       are the width and the height of the sheet of paper. The sheet of paper
       is represented only by a frame on the screen. It does not affect the
       drawing of the DVIfile.

       When pagemark and screenmark are shown, the argument (*) sets the
       boarder of the page to the rectangle described by screenmark and
       pagemark.  When they are hidden, (*) takes the command-line-options
       resp. defaults -h,-v and -p.


HALF-HYPER ********************************************************************

tmview does some of the fancy hyper-tex things. I talk about HALF-hyper-tex,
because tmview follows only links which point to somewhere within the
currently visited dvi-file. So there is no connection to the net or so. But
you might find it usefull (when editing a major project) to view an equation
number this-and-that by clicking on this-and-that whereever the text refers to
that equation.  For information about hyper-tex, related macropackages and
fully compatible viewer scan the net ...

<k> Show/hide half-hyper-mark
<TAB> Goto next href
<RET> Follow current href, if any 


MISC **************************************************************************

<s> Search for text

      You will be asked for the text-string to be searched. You may enter a
      regular expression describing that string, that includes especially just
      to enter the string as it is.

      tmview will take the entire DVIfile as one huge text-string and then
      search for the next substring, fitting the regular expression you've
      enterd. Thereby "next" is ment with respect to the current page.

      So far this sounds quite easy, but there are some ugly details, based on
      the fact, that a DVIfile contains information on how to draw a bitmap
      representing your text. It does not contain information about from what
      characters in which order your text is made up. Even the PKfiles used to
      draw your text consist only of lots of glyphs but no character-codes,
      like ASCII or so. Building a huge text-string from a DVIfile is some
      kind of guessing.

      Fisrt: What kind of huge-string is build from the DVIfile?

      This string will consists of the letters <A> ... <Z>, <a> ... <z>, the
      accent <"> and the digits <0> ... <9>. It does NOT contain anything
      else, like <space>, <ret> or <->: 
      Whenbeingprinteditwouldlooklikethisnotreadableatall.  
      Taking the DVIfile as huge string allows you to find all locations of a
      sub-string, say "commandlineoptions", even those that are seperarted by
      linebreaks (and hyphens) or pagebreaks. In turn, there is no chance to
      find all those locations, where "commandlineoptions" is seperated by a
      hypenation.  To keep tmview from being confused by headings, there is
      another rule for building up the huge text-string: any glyph outside the
      printable area (see <p>) is ignored. So you may setup the printable area
      to ignore headings when searching.

      Second: How is the huge-string build up?

      To translate the list of glyphs found in the DVIfile to a text-string,
      the tfm-files are asked for the encoding-scheme. This does work with
      dc-fonts and cm-fonts, since the following encoding-scheme names are
      accepted: "ASCII", "TeX text", "TeX math italic", "TeX math symbols",
      "TeX typewriter text", "Extended TeX Font Encoding - Latin", "Adobe
      StandardEncoding".  
      The alphanumerics <A>...<Z>, <a>...<z>, <0> ... <9> are copied one by
      one to the huge text-string. Glyphs that "look like" a simple
      alphanumeric will be taken as that one it looks like.  So the Tex input
      '\c o', producing an 'o'-with-an-cedilla-accent, will be represented as
      a simple (o) in the text-string. This rule also works for all kind of
      ligatures. The TeX input 'ffl\AE' will be represented by (fflAE).  Any
      accent ON TOP of a glyph will be translated to a ("), preceding whatever
      the glyph without that accent would be translated to.  The TeX input
      '\"a' producing the german umlaut 'a'-with-two-dots-on-top, will be
      found as ("a) in the generated text-string.  The TeX input '\aa'
      producing the scandinavian 'a'-with-circle-on-top will be found as ("a)
      too. Any other glyphs are ignored.

      Third: In what does the above result?

      Visiting english documents, say manuals to some computer related stuff
      like elisp.dvi, searching for keywords works fine. Searching in
      documents in which extensive use of accents and funny characters is made
      works too, but requires some luck or/and experiance in how TeX acts on
      such things.


      Example: 

      Take the file story.tex from the TeXbook, chapter 6, page 24.  It
      contains the line
              galaxy called \"O\"o\c c,
      The text-string build from the corresponding story.dvi will therfore 
      contain
              galaxycalled"O"oc
      You may search for ...    getting as result ...
      galaxy                    found
      galaxycalled              found
      galaxy called             not found
      d"                        found
      galaxy.*"O"oc             found
      Ooc                       not found
      
  
<r> Re-read current DVIfile and re-draw screen

      When the argument <*> is given, re-read the font-files too.


<d> Load/kill DVIfile

      After typing <d> you may select between <l> to load a DVIfile and <k> to       kill a DVIfile. 
      
      Loading a DVIFile: 

      tmview will look for a file-bookmark belonging to that file. If there is
      one, it becomes the current file-bookmark. The DVIfile will be shown as
      left, and any defined manual-bookmarks are accessable by <w>. When
      loading a DVIfile for the first time, a new file-bookmark will be
      generated.  This will be setup with default values from the command-line
      options and won't contain any manual-bookmarks.

      Killing a DVIfile:
      To kill a DVIfile means to kill it's file-bookmark and any related
      manual-bookmark. Killing a DVIfile won't hurt the file itself.  You
      don't have to kill a DVIfile just to load another one.
     
<q> Quit tmview 
    
      When quitting, a startup-file will be written. When running tmview
      next time, you will find almost everything as you left it.


*******************************************************************************
THE STARTUP-FILE***************************************************************
*******************************************************************************

The following explains the contents of the startup-file used to setup
tmview. When not set read-only, tmview writes down a startup-file every time
you quit tmview. Therefore the list of bookmarks resp. visited files will keep
growing, solong as you don't explicitly kill them with <d><k>.  Since also
command-line options are kept in the startup-file you are not able to reset
any nonsense by quiting/restarting tmview.  To prevent such a behavioral,
simply set the startup-file read-only.  But there are two good reasons to make
tmview remember all the above mentioned information:


1. You are working on a major TeX project, creating a huge document. 
This will take some days, and you still turn the system off sometimes. 
You want tmview to remember all the bookmarks you've set in your document,
the abnormal papersize, the paths to special fonts used nowhere else, a.s.o. 
Since that kind of information is tied to your document, call tmview
with "-sMyDoc.tmv". So the tmview related settings on that project
will be kept in "MyDoc.tmv" and won't disturb your standard startup-file
in "~/.dvisvga". 

2. You are using tmview as a viewer of serveral manuals.
This is similar to the above. Every once and a while you should save the
current startup-file, to prevent loosing the valueable information
(=bookmarks) by accident.

However, as mentioned above, you might one day end up in a somehow ugly large
or even corrupted startup-file. The easy way out is to remove the
startup-file, call tmview with suitable standard-options and thereby get a new
startup-file.  For the case you want to copy some of the information contained
in the old startup-file to the new one, the internal structure of the
startup-file will be explained by the following example startup-file:

# **** this is an example startupfile of tmview *****
#
#
# basic rules:
# 1. all charakters from a '#' to the end of a line are ignored.
# 2. spaces, tabs and newlines act as seperators.
# 3. a symbol starts with a letter, consists of letters and digits,
#    and ends at the next seperator.
# 4. a string starts just after a '"' end ends just befor the next '"'.
#    exeption: to get a string containing a '"' type  '""'
# 5. a number starts with '+' '-' '.' or a digit and must look like
#    numbers look like, i.e. no two '.' and so on.
# 6. a value is a list of strings and/or numbers seperated by seperators.
# 
# The purpose of the startup-file is to set symbols to values. This is done
# by typing the symbol followed by a value (seperated by ... guess ...)  
# Therfore a number of valid symbols are defined, each expecting some
# kind of value. If a valid symbol is not set in the startup-file, a
# default value is used. You may therefore remove any suspect seting from you
# your startup-file.

# most important: the symbol sufp
# Set the startupfile. Since this is read from a startup-file, and 
# that one has to be known therefore, setting sufp sets the name of the file
# to write startup information when quitting tmview.
# Not to set sufp prevents writing a startup-file at all.
# Setting sufp to some other file prevents overwriting the current 
# startup-file, while still saving newly generated bookmarks a.s.o.
  
sufp "/home/thomas/.dvisvga"

# now set the commandline-options ...

hoff  25.40  # -h(hoff) 
voff  25.40  # -v(voff)
papx 210.00  # -p(papx)x(papy)
papy 297.00
xres 300     # -r(xres)x(yres)
yres 300
tfmp "./:/usr/lib/texmf/fonts/tfm//:/usr/lib/texmf/fonts//"     # -t(tfmp)
vffp "./:/usr/lib/texmf/fonts/vf//:/usr/lib/texmf/fonts//"      # -g(vf)
ffor "@N.@Mpk"                                                  # -n(ffor)
fntp "./:/usr/lib/texmf/fonts/tmp/pk//:/usr/lib/texmf/fonts//"  # -f(fntp)
disx 640     # -d(disx)x(disy)
disy 480
nmag 0       # -m(nmag)
lrnd   4.00  # -k(lrnd);(rrnd);(ornd);(urnd)
rrnd   4.00
ornd   4.00
urnd  12.00


#other defaults 

verb 0      # 'verb 1' makes tmview a bit more verbose


# file-bookmarks and manual-bookmarks
#
# The symbol fmk starts the definition of a file-bookmark. 
# It is followed by the name of the file as a string and lots
# of other values. Manual-bookmarks, belonging to a file, have to follow
# immediatly, starting with the symbol bmk. This results in blocks
# starting with fmk and ending just before the next fmk, containing
# all the information related one DVIfile.
# 
# You may either delete some of the lines starting with bmk to get rid
# of those bookmarks. Or you may delete all the lines beloning to one
# file-bookmark, including the manual-bookmarks that follow. Be shure,
# not to leave some of the related manual-bookmarks, when deleting
# a file-bookmark.
# You may either copy a file-bookmark only, copy or a file-bookmark together
# with the related manual-bookmarks from another startup-file.
 
# one block, containing 3 manual-bookmarks
fmk "/home/thomas/cc/tmview/asmith.dvi" 210.00 297.00  25.40  25.40 105.00 148.50   4.00   4.00   4.00  12.00 0 1 -7 389  3.000 
bmk 3 1 -7 517  3.000 
bmk 1 1 -7 261  3.000 
bmk 2 1 -7 389  3.000 


# another block, not containing manual-bookmarks at all.
fmk "/home/thomas/cc/tmview/tmview/dort.dvi" 186.10 276.78  17.19   8.89 186.10 261.37   4.00   4.00   4.00  12.00 0 3 -163 -120  7.000 

# end of example startup-file



*******************************************************************************
COPYING************************************************************************
*******************************************************************************

The code of tmview uses some rather basic ideas stolen from xdvi. This
includes some few lines of code just copied. The author of xdvi is Eric
Cooper. In a similar kind, tmview's code depends on some lines of dvidjc,
written by Wolfgang R. Mller. The hyper-TeX related parts are taken from
xhdvi, written by Arthur Smith. The sources of the above can be found on the
CTAN. So by having mentioned the authors here, and giving a reference how to
get the original sources, this should not be a violation of their copyrights.
 
As far as I am concerned, tmview may be modified or distributed without any
restrictions. tmview is distributed in the hope that it will be useful, but
without any warranty.

*******************************************************************************
End of README******************************************************************
*******************************************************************************