license violation in tetex-texmf-3.0 fixed
Petr Olsak has notified me today, that tetex-texmf-3.0.tar.gz contains
three files which have been modified in a way that violates his license.
The modification had been done by me (as I wrongly assumed that the
license was GPL; in fact the license was GPL with some additional
restrictions).
So, I just have updated these three files (reverted to the original
version) and replaced the broken tetex-texmf-3.0.tar.gz by the fixed
tetex-texmf-3.0po.tar.gz (I had announced that I won't make a new release.
Hey, this does not really count :-) I will take a small ammount of time
until CTAN will see the update.
From the Changelog:
Do Mai 25 18:28:58 CEST 2006
* fixed a copyright violation, caused by distributing modified
files of cstex / cslatex under their original name. The fix
reverts to the original files. Affected files:
tex/cslatex/cslatex.ini
tex/csplain/cseplain.ini
tex/csplain/csplain.ini
This update just fixes the copyright violation. I cannot recommend to
download this update. The version of cstex / cslatex contained in
teTeX-3.0 (both tetex-texmf-3.0.tar.gz and tetex-texmf-3.0po.tar.gz)
is severely broken (e.g. pdfcslatex produces dvi by default). So, if you
are interesed in using cstex / cslatex, better download and install the
latest version.
The official web site of cstex, http://math.feld.cvut.cz/olsak/cstex/
refers to ftp://math.feld.cvut.cz/pub/cstex/ for downloading the latest
version. I see that some version exists on CTAN, but I am not sure if
CTAN carries the latest version.
All the best,
Thomas
PS: the changes are so small, that I append the full diff to this mail.
diff -ru tetex-texmf/ChangeLog tetex-texmf-po/ChangeLog
--- tetex-texmf/ChangeLog 2005-02-06 20:05:00.0 +0100
+++ tetex-texmf-po/ChangeLog2006-05-25 19:13:19.0 +0200
@@ -1,3 +1,12 @@
+--- teTeX-3.0 po
+Do Mai 25 18:28:58 CEST 2006
+ * fixed a copyright violation, caused by distributing modified
+ files of cstex / cslatex under their original name. The fix
+ reverts to the original files. Affected files:
+ tex/cslatex/cslatex.ini
+ tex/csplain/cseplain.ini
+ tex/csplain/csplain.ini
+
--- teTeX-3.0
So Feb 6 13:05:44 CET 2005
* update fpl fonts
diff -ru tetex-texmf/release-tetex-texmf.txt
tetex-texmf-po/release-tetex-texmf.txt
--- tetex-texmf/release-tetex-texmf.txt 2005-02-06 20:04:02.0 +0100
+++ tetex-texmf-po/release-tetex-texmf.txt 2006-05-25 18:31:46.0
+0200
@@ -1 +1 @@
-3.0
+3.0po
diff -ru tetex-texmf/tex/cslatex/cslatex.ini
tetex-texmf-po/tex/cslatex/cslatex.ini
--- tetex-texmf/tex/cslatex/cslatex.ini 2004-03-26 20:09:41.0 +0100
+++ tetex-texmf-po/tex/cslatex/cslatex.ini 1998-12-04 18:47:38.0
+0100
@@ -5,15 +5,6 @@
%% tex -ini \let\enc=w \input cslatex.ini
%% see the csplain.doc in csplain.tar/zip for more details
-\ifx\pdfoutput\undefined
-\else
- \ifx\pdfoutput\relax
- \else
-\input pdftexconfig
-\pdfoutput=0
- \fi
-\fi
-
\ifx\enc\undefined \else %% Reencoding by encTeX is needed
\begingroup
\catcode`\{=1 \catcode`\}=2 \catcode`\#=6
diff -ru tetex-texmf/tex/csplain/cseplain.ini
tetex-texmf-po/tex/csplain/cseplain.ini
--- tetex-texmf/tex/csplain/cseplain.ini2004-03-26 20:09:59.0
+0100
+++ tetex-texmf-po/tex/csplain/cseplain.ini 1998-12-02 14:50:30.0
+0100
@@ -1,13 +1,4 @@
% Hacked from csplain.ini by Jan Yenya Kasprzak [EMAIL PROTECTED]
-\ifx\pdfoutput\undefined
-\else
- \ifx\pdfoutput\relax
- \else
-\input pdftexconfig
-\pdfoutput=0
- \fi
-\fi
-
\input csfonts % re-defines primitive \font
\input plain% format Plain
\restorefont% original meaning of primitive \font
diff -ru tetex-texmf/tex/csplain/csplain.ini
tetex-texmf-po/tex/csplain/csplain.ini
--- tetex-texmf/tex/csplain/csplain.ini 2004-03-26 20:09:47.0 +0100
+++ tetex-texmf-po/tex/csplain/csplain.ini 2000-02-25 11:49:18.0
+0100
@@ -1,12 +1,3 @@
-\ifx\pdfoutput\undefined
-\else
- \ifx\pdfoutput\relax
- \else
-\input pdftexconfig
-\pdfoutput=0
- \fi
-\fi
-
\input csfonts % re-defines primitive \font
\input plain% format Plain
\restorefont% original meaning of primitive \font
Re: texconfig action and site-wide changes
Hi Frank, It seems to me that texconfig-sys init should call fmtutil-sys and updmap-sys, too, not fmtutil and updmap plain - I didn't check for texlinks.] That is wrong, just try texconfig-sys formats change some bits in the config file (-goes via fmtutil --edit to TEXMFSYSCONFIG) and the changed/added formats to via fmtutil --byfmt to TEXMFSYSVAR). The environment manipulation done in texconfig-sys has of course not only some effect on the texconfig script that it calls, but also to all other subprocesses. I have tried to understand how texconfig works if a user invokes it. As far as I can see, for every configuration file they touch, a copy is generated in TEXMFCONFIG (which is $HOME/.texmf-config by default). Right. This seems to have the (probably unwanted effect) that the user is thus cut-off from site-wide changes. The same is true if some user puts a custom copy of koma-script into his $TEXMFHOME. If a user changes a configuration file $cfile for the first time, the changed file after check_out is not only copied or cat'ed to $TEXMFCONFIG/$relDir/$cfile, but additionally a diff or the change regex Well, this sounds like a complicated solution (different config files would need different kind of updates) with a questionable effect. Thomas
teTeX-3.0
teTeX-3.0 has appeared on the Cambridge CTAN node. I will send a more official announce as soon as all CTAN sites are updated. But, I thought that you might want to grab the brand new release right now. The xdelta files are available at ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/ and the full tarballs at ftp://cam.ctan.org/tex-archive/systems/unix/teTeX/3.0/distrib/ The teTeX homepage is updated, too: http://tug.org/teTeX/ Before I give you the list of changes, I'd like to say a big big *THANK YOU* to everyone who has helped to get this out in this (hopefully) good quality. The feedback that I have received has been extremely helpful and some people have constantly contributed a significant amount of their time for helping me debugging and solving several problems! Concerning 2.99.13, I have only received positive feedback (9 messages) and I hope that I didn't mess it up with one of my last minute changes. Thanks, Thomas Changes in the src area: * texconfig: add font ro font rw * Update TETEXDOC for texconfig additions; update kpathsea docs for texmfvar feature. * mktexdir: continue with next dir if mkdir fails * update ChangeLog, NEWS, release date of web2c-7.5.4 in web2c / kpathsea (from Olaf) * changes to texmf.in: pdfxtex paths and comments * update dvipsk/writet1.c by the new version in pdftex Changes in the texmf area: * update fpl fonts * update doc/programs from tetex-src tarball * update antt fonts * update fontinst (1.928 - 1.929) * update the pdftex manual * update from Heiko: graphics.cfg color.cfg
teTeX-3.0
I am pleased to announce release 3.0 of the teTeX distribution. It was
promised for sometime in 2004 and I have only slightly missed that
goal. Thanks to everone who has helped to get this out!
The teTeX homepage (http://tug.org/teTeX/) now has some files of the
source distribution (e.g. the INSTALL document) and the complete doc tree
of the texmf distribution (http://tug.org/teTeX/tetex-texmfdist/doc/)
online, so you can see in detail what you get before you download /
install.
February 2005, Thomas Esser
===
teTeX 3.0
===
This is the announce of teTeX-3.0, a TeX distribution for UNIX
compatible systems.
===
Main features of teTeX-3.0:
===
- teTeX is free software (see http://www.fsf.org/philosophy/free-sw.html)
- included programs: web2c 7.5.4, xdvik 22.84.8, dvipsk 5.95a, texinfo
4.8, pdfetex 1.21a (based on TeX-3.141592 and eTeX-2.2), Omega 1.23.2.3,
dvipng 1.5, Aleph 3.141592-1.15-2.1-0.0-rc4 (combines a stable omega
version with features of eTeX), dviljk 2.6p2, ps2pk 1.5, makeindex 2.14,
texi2dvi-1.76, texconfig, fmtutil, updmap, texdoc, texdoctk v0.6.0a
- TDS (TeX Directory Structure) rev. 1.1 compliant support tree
with fonts / macros / documentation: 226 MB, 13000 files
- updated texmf tree with many updated updated / added packages. Most
notably: LaTeX2e 2003/12/01, ConTeXt version 2005.01.24, Latin Modern
fonts, beamer, koma-script and memoir package
- easy to install and to customize, even for a multi-platform setup
- ready for producing resolution independent (bitmap free) postscript or
pdf documents (including thumbnails, hyperlinks and bookmarks)
For release notes, please look at NEWS or texk/tetex/doc/TETEXDOC.pdf
inside the tetex-src tarball. These files are also available online:
http://tug.org/teTeX/tetex-src/NEWS
http://tug.org/teTeX/tetex-texmfmain/doc/tetex/TETEXDOC.pdf
===
Legal notice / license
===
teTeX is free software. Individual files and sub-packages are copyright
by different authors, but I have taken care that everything is free
software as defined by the free software foundation:
http://www.gnu.org/philosophy/free-sw.html
===
The files / installation
===
You can get teTeX-3.0 from the following server:
ftp://cam.ctan.org/tex-archive/systems/unix/teTeX/3.0/distrib/ (UK)
Other servers will follow shortly:
ftp://dante.ctan.org/tex-archive/systems/unix/teTeX/3.0/distrib/ (GER)
ftp://tug.ctan.org/tex-archive/systems/unix/teTeX/3.0/distrib/ (US)
A full list of mirrors is available here:
ftp://cam.ctan.org/tex-archive/CTAN.sites
From these servers, you need the following files for teTeX-3.0:
sources of the programs (required):
== tetex-src-3.0.tar.gz
support tree with fonts / macros (required):
== tetex-texmf-3.0.tar.gz
sources for parts of the texmf tarball (optional):
== tetex-texmfsrc-3.0.tar.gz
The installation instructions are included in the tetex-src.tar.gz
archive as files INSTALL (complete instructions) and QuickInstall
(short version) and also available online:
http://tug.org/teTeX/tetex-src/INSTALL
http://tug.org/teTeX/tetex-src/QuickInstall
===
MacOS X version
===
I don't maintain this port myself, but you can get it here:
MacOS X (darwin) http://www.rna.nl/tex.html
===
Documentation / getting started
===
teTeX contains a lot of documentation about the packages it includes.
For a start, please look at
$prefix/share/texmf/doc/tetex/TETEXDOC.{ps,pdf,dvi}
and
$prefix/share/texmf-dist/doc/index.html
prefix means the directory in which you have installed teTeX according to
INSTALL or QuickInstall.
Online versions are also available:
http://tug.org/teTeX/tetex-texmfmain/doc/tetex/TETEXDOC.pdf
http://tug.org/teTeX/tetex-texmfdist/doc/index.html
===
De-support of repacked and binary distribution
===
Starting with teTeX-2.0, I have stopped to make binary releases or
sub-packges
Re: 2.99.12 (Irix failure)
Either would work for me... But note that ttf2afm.c is already using xstrdup right over there as well instead of testing for it in the system libraries - That is right. We can't make a fix just for Irix. Using xstrdup is the rigth way of fixing it. Thomas
Re: 2.99.11: yet another release candidate
For Cygwin, I have been using using the following commands during postinstall unset MFINPUTS TEXINPUTS TEXMF TEXMFCNF TFMFONTS VARTEXFONTS texconfig rehash texconfig confall texconfig rehash texconfig init # default dvips output to file texconfig dvips printcmd - to create the initial formats. Can these just be replaced by texconfig-sys, Yes, you should use texconfig-sys. It should be sufficient to run mktexlsr texconfig-sys init texconfig-sys dvips printcmd - and will that then update as before to the main tree? No, this will not update $TEXMFMAIN. The map files / formats will be written to TEXMFSYSVAR. (I'll test with .12 if I get the chance.) Please, check if the settings for TEXMFVAR / TEXMFCONFIG make sense on Cygwin. Thomas
Re: 2.99.12
I haven't tested .11, with .12 I found that ttf2afm.c blondely includes libgen.h without check. Thanks. pdftex-1.21a uses xbasename instead of basename, but still includes libgen.h. So, I just have removed that #include libgen.h. Thomas
2.99.12
Well, this time, we don't have many changes (which is a good thing!).
Martin wants to make the pdftex-1.20c release really soon, and this
might give us the opportunity to check if the current pdftex beta has
any problems (well, we really won't have much time to test this).
Similar, texi2html-1.76 will be released, soon. The current beta should
work with more perl versions. ANd what is more important: even if your
perl is not compatible with texi2html or if you don't have perl at all:
the teTeX build will no longer break.
I think that the final 3.0 release can be made this weekend; if no
serious problems pop up.
As usual: please, send me feedback, even if it all just works :-)
Files (to appear soon):
ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-2.99.12.20050203-beta.tar.gz
ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmf-2.99.12.20050203-beta.tar.gz
ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-2.99.12.20050203-beta.tar.gz
or
ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-src-2.99.12.20050203-beta.tar.gz
ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmf-2.99.12.20050203-beta.tar.gz
ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-2.99.12.20050203-beta.tar.gz
Changes in the src area:
* update pdftex (1.20c-beta-20050121 - 1.20c-beta-20050203)
* update texi2html (1.74 - 1.75; 1.75 is a test release!)
* portability fix to {texconfig,fmtutil,updmap}-sys
* adjust line ends in dvipsk/search.c, dvipsk/tfmload.c
* enhance the choose a mode text in texconfig-dialog
Changes in the texmf area:
* update 2 context files (from Hans): spec-dpx.tex spec-fdf.tex
* update doc/latex/fp/readme.fp (now defines lppl as license
for the package)
* update context (2005.01.24 - 2005.01.31)
* update doc/index.php / doc/index.html
* update texinfo.tex
Happy testing + have fun!
Thomas
Re: 2.99.12
Hi Robin, pulled and pushed, as always :-} Great, thanks. Thomas
Re: 2.99.11: Compilation under irix broken
Hi Ezra, thanks for your report. I suspect that the version of perl is at fault... Looking at the script it seems that 5.004 is all that is required... I am in contact with the texi2html people and they are very responsive and helpful. They have already provided a version for me that no longer fails if perl does not exist or if it does not work for texi2html. They might even fix texi2html to work with older perl versions (the texi2html docs claim that any version of perl5 works, which is currently not true; but that's not that important for me; the main point is that the teTeX build/install no longer fails). Thomas The build log... making all in utils/texi2html Making all in . Making all in doc TEXINPUTS=.::$TEXINPUTS \ /usr/sbin/perl ../texi2html -V texi2html.texi Bareword our not allowed while strict subs in use at ../texi2html line 2422.Bareword old_style_map not allowed while strict subs in use at ../texi2html line 2422. Operator or semicolon missing before %old_style_map at ../texi2html line 2422. Ambiguous use of % resolved as operator % at ../texi2html line 2422. Global symbol %old_style_map requires explicit package name at ../texi2html line 2850. Bareword our not allowed while strict subs in use at ../texi2html line 3083.Bareword old_style_map_pre not allowed while strict subs in use at ../texi2html line 3083. Operator or semicolon missing before %old_style_map_pre at ../texi2html line 3083. Ambiguous use of % resolved as operator % at ../texi2html line 3083. Global symbol %old_style_map_pre requires explicit package name at ../texi2html line 3084. BEGIN not safe after errors--compilation aborted at ../texi2html line 5827. *** Error code 255 (bu21) *** Error code 1 (bu21) *** Error code 1 (bu21)
Re: 2.99.11: yet another release candidate
Hi Erik,
thanks for your testing and feedback!
I am not sure where generated font-files are best placed, but my
understanding is that $TETEXDIR/share/texmf should hold any
automatically generated stuff like format files and font files.
The default is to have the font files in /var/tmp/texfonts. If you want
to use the texmfmain tree for that, you should:
texconfig-sys font vardir $TETEXDIR/share/texmf/fonts
mkdir $TETEXDIR/share/texmf/fonts
chmod 1777 $TETEXDIR/share/texmf/fonts
That way, the fonts directory (and any automatically created subdirectory)
will be writable by everybody.
I might add something like this to the documantation or add support for
it to texconfig.
Since I have failed to see the reason for a $TETEXDIR/share/texmf-var
tree when I already have the $TETEXDIR/share/texmf tree, I have omitted
the texmf-var tree.
Still, you *have* to set TEXMFSYSVAR and TEXMFSYSCONFIG in your texmf.cnf
file. You can assign the same value taht TEXMFMAIN has, but you have to
set it.
I haev also always wondered how to manage write permission for the font
directories.
It used to be simple. Now, we have $TEXMFMAIN/web2c/mktexdir which
uses kpsestat. To some extend, permissions of the parent directory are
inherited (and a sticky bit might be added).
feels highly unsatisfactory. What is the right way to do it?
Set the right permissions to the parent directory.
However, if I then do 'fmtutil-sys --all', format files still get
generated in ~/.texmf-var, not in $TETEXDIR/share/texmf as I wanted (or
$TETEXDIR/share/texmf-var as I had expected)
I see that /bin/sh on solaris does not put a=b into the environment if
exec is used in the following way:
/bin/sh -c 'a=b exec env'
I have not expected this and I guess that on most systems with a standard
bourne shell this makes all the -sys variants useless. Sorry! (I am
grateful that this was detected that early!)
Try to use
TEXMFVAR=$v; TEXMFCONFIG=$c; export TEXMFVAR TEXMFCONFIG; exec fmtutil
${1+$@}
instead of
TEXMFVAR=$v TEXMFCONFIG=$c exec fmtutil ${1+$@}
(same change for texconfig-sys and updmap.sys).
Can someone enlighten me how to set up the environment.
It is not your fault!
Also on a final note, I have the cm-super fonts installed in the local
texmf-tree and would like to enable them. Previously I have edited
updmap.cfg by hand but now I tried to use the updmap-sys script like this:
updmap-sys --edit
does what you want.
...
updmap-sys --enable Map=cm-super-x2.map
This does not seem to work as expected.
It does not? How does it fail? What is the error/problem?
Thomas
Re: Info about texconfig
[Cc: list changed to tetex-pretest, because we are discussing the 2.99.10 version] Dear Fabio, thanks a lot for your feedback about texconfig. I think that some points are easy to answer (questions about modes), others are really week points of the current implementation (see below). For example, in the Mode setup there are hundreds of choices but my guess is that, all in all, the only thing that is really important is the resolution of the printer, isn't it? No, but thanks for your questions. I justhave changed the text to make it more clear: Chosse a mode to be used when metafont generates font bitmaps. The resolution is the most important point, but there might be differences between modes of the same resolution. See the comments in the file modes.mf for more details. Or is there a difference (which?) between epsonsq (360 dpi) and epstylus (360 dpi)? Just look up the definitions in the modes.mf file... If I print indifferently on 2 printer (epson stylus photo 1200 and hp4050) which mode should I choose? You can set up individual modes in the dvips section. Add 2 printers and choose the optimal mode for each of them. If you have no idea which mode to choose, read the comments in the modes.mf file. Moreover, what does the mode section sets up? It is all about bitmap fonts. Possibly, you don't use them much because teTeX contains so much type 1 fonts. Even xdvi uses them, so bitmap fonts are only generated for fonts which have no type 1 equivalent. If you choose mode from the main menu, the following will be set up: - mode to be used by mktextfm - mode + dpi setting in dvips's config.ps - resolution setting for pdftex's pdftexconfig.tex I mean: if I want a quick mode only for viewing on the screen with xdvi Stefan Ulrich has suggested to me to keep a fixed setting of mode=ljfour and dpi=600 for xdvik, independend from texconfig. 600dpi are good for today's 100dpi displays, because xdvik uses antialiasing. Another point which, IMHO, is confusing is that there is no way to see the actual settings (the conf voice doesn't say a lot about actual settings...) and you can guess even less about which printers are defined and how. That is a very true point. Something could be done better, other things are deficiencies of the dialog program that is used. Today's design of texconfig allows to replace the user interface. Someone could write a perl/tk interface or perl/qt or soemthing different. Instead of making a pleasant user interface, I have significally enhanced the backend: the command line interface of texconfig. It is not fast and does nothing more than what the command line says. The whole interface that you don't like (and believe me: I don't like it very much, too) is texconfig-dialog. One last question: there will be major differences between 2.99.10 and the forthcoming version 3 of tetex? On my system (linux mandrake 9.0) 2.99.10 seems to work very well, so maybe it is not worth, for me, to recompile again. Yes. There have been lots of changes for 2.99.11 / 2.99.11b. I suggest that you upgrade at least when 3.0 is out. Thomas
Re: texshow and kpsewhich
to get texshow working with teTeX-2.99.10.20050123-beta, I needed to add CONTEXTINPUTS = .;$TEXMF/tex/context// to texmf.cnf No, that sounds like texshow is using the wrong search path --format='other ... files' --progname=context where it should better have been --format=.tex --progname=context Please, report this to Hans Hagen. It would be great to see a fix in one of the next days. Thomas
Re: 2.99.10: another release candidate
the patch below includes xdvi.h (which then includes strings.h) _before_ Xm.h which avoid this trouble. Looks good, thanks! Thomas
Re: 2.99.10: another release candidate
ln -s dvips.1 /opt/local/share/man/man1/odvips.1 ln: `/opt/local/share/man/man1/odvips.1': File exists Argh... The line ln -s dvips.$(manext) $(man1dir)/odvips.$(manext) should be removed from texk/tetex/Makefile.in. Thanks for reporting this. /opt/local/share/man/man8, where the sources of all these links should be, and only fmutil.8 was there. Actually, all of the above links up to No, section 8 does not mean sources of manpages. It means administrative commands. man8. IThis is probably something trivial (and most likely my fault), but I am totally befuddled. No, my fault. Thomas
Re: 2.99.9: 3rd (and last?) release candidate
This is why dvipng.h does not read stdint.h. Am I wrong in following this advice? No, it seems just right not to include stdint.h in dvipng.h. On the other hand, I wonder why stdint.h should be included in the autoconf test... That seems wrong to me. (Actually, the dvipng distribution uses autoconf 2.5x, but tetex does not. And I don't suppose Thomas wants to change this close to release.) Right, I am not going to change this now. Thomas
Re: 2.99.8: updmap
It works well on Vine Linux 2.6/3.1 (bases on redhat). Thanks for testing and thanks for your report. On bash-2.05 it is ok, but on bash-2.05b Hm... Two questions. What is the output of type ls on your two different systems (bash-2.05 / bash-2.05b)? Does the original script work, if you replace ls -l $outputFiles | verboseCat by \ls -l $outputFiles | verboseCat - OLDIFS=$IFS; IFS=$newline + # OLDIFS=$IFS; IFS=$newline ls -l $outputFiles | verboseCat - IFS=$OLDIFS + # IFS=$OLDIFS verboseMsg That would cause other problems, however. The updmap script is very clean about whitespace in directory names and not setting IFS would break this. Try: TEXMFVAR=$HOME/.texmf var updmap (BTW: I just found that mktexlsr needs a whitespace related fix :-( Thomas
Re: 2.99.7 ConTeXt, URW-Fonts
i'm kind of puzzled beause i thought that 8r was as frozen as can be (i must admit that i never use that encoding explicitly); There have been several changes to 8r.enc in the past. Walter took over the maintenance (the file is now part of psnfss) at a time where different versions have been floating around. Have a look at CTAN:macros/latex/required/psnfss/changes.txt doesn't such a change also relate to the presense of glyphs in the pfb's? Yes and I think that this is the main reason for the changes. E.g. many fonts (including the urw35) now have the Euro, so it was added. Thomas
Re: 2.99.7: first rc of teTeX-3.0
On Tue, Dec 28, 2004 at 08:36:21PM +0300, Vladimir Volovich wrote: - sparc/Linux, GNU gcc-3.3.5, GNU flex 2.5.31, GNU bison 1.875d, GNU make 3.80 Thanks. Though, there is a bug in pdfetex which makes it impossible to use hyperlinks in right-to-left mode. It would be very nice to have that fixed before the release... :) Yes, I have seen that bug report. If we get a tested fix for it in one of the next days, that might get in. Otherwise, we have to live with that bug. Thomas
Re: 2.99.7: first rc of teTeX-3.0
Ezra, thanks a lot for this early feedback (very much appreciated)! Okay - everything compiles under irix 6.5.20m with the base compiler and make. Good. However, texconfig does not run interactively... I believe that tcdialog is having issues - it just exits - the problem is the TERM/TERMINFO settings. Your analysis is right, DIALOG_TERMINFO should point to $TEXMFMAIN/texconfig. The old script has done this and I have forgotten to copy that part of the code. I'll fix that. I did not notice this bug, because I am using Linux and this has a special case to use /usr/bin/dialog instead of tcdialog. Separate wish list... It would be useful to have texconfig paper return the current setting... Hm... Sounds good. (The same is certainly true for a few other settings as well.). I'll have a look and see how much work this is. Also - texconfig help indicates that one could configure the paper size seoarately for xdvi, dvipdfm. Are these stll valid? The menu access in texconfig does not allow these to be individually set. Here, the batch mode has more capabilities than the interactive mode. I thought it might be better to restrict the interactive tool to the most basic tasks. That way, it is easier to find the important things. Initial tests show that font creation is going into the share/texmf tree now. That is intentional, because TEXMFVAR points there. Font creation used to go to the tree of the font source (if that was writable), but to fit into the general concept, I have changed that (it can be undone by removing texmfvar from the file mktex.cnf). Is there any problem with this approach? Thomas
Re: 2.99.7: first rc of teTeX-3.0
On Mon, Dec 27, 2004 at 09:39:24AM +0100, Thomas Esser wrote:
Your analysis is right, DIALOG_TERMINFO should point to
$TEXMFMAIN/texconfig. The old script has done this and I have forgotten
to copy that part of the code. I'll fix that.
A quick fix could be like this:
*** texconfig-dialog.broken2004-12-25 21:36:41.0 +0100
--- texconfig-dialog 2004-12-27 10:54:40.047835169 +0100
***
*** 428,433
--- 428,436
: ${TEXMFMAIN=`kpsewhich -var-value=TEXMFMAIN`}
export TEXMFCONFIG TEXMFVAR TEXMFMAIN
+ : ${DIALOG_TERMINFO=$TEXMFMAIN/texconfig}
+ : ${DIALOG_TERM=generic}
+
setupTmpDir
echo $progname: started `date` $log
findDialog
Thomas
Re: 2.99.7 ConTeXt, URW-Fonts
On Mon, Dec 27, 2004 at 08:18:50PM +0100, Hartmut Henkel wrote: texmf/fonts/enc/dvips/txfonts/tx8r.enc texmf/fonts/tfm/urw35vf/*/u*.tfm texmf/fonts/vf/urw35vf/*/u*.vf Well, there is some long discussion behind this problem. I think that the URW fonts are well enough supported by the Adobe metrics, so I don't like to add these special metrics to teTeX. It would be nice to be able to typeset the pdftex manual with the basic set of support files provided by teTeX (which included cont-tmf.zip). Thomas
Re: 2.99.7 ConTeXt, URW-Fonts
On Tue, Dec 28, 2004 at 05:10:53AM +0100, Reinhard Kotucha wrote: There is no difference between tx8r.enc and 8r.enc and this is the reason tx8r.enc has been removed from teTeX. This used to be true, but now it is even worse. tx8r.enc and 8r.enc differ (because 8r.enc was changed and tx8r.enc remained unchanged): After bringing both into a normalized form, I get: $ diff tx8r.enc 8r.enc 24,31c24,31 /.notdef /.notdef /.notdef /.notdef /.notdef /.notdef /.notdef /.notdef --- /notequal /infinity /lessequal /greaterequal /partialdiff /summation /product /pi 130,131c130,131 /.notdef /.notdef --- /Euro /integral 143,145c143,145 /.notdef /.notdef /.notdef --- /Omega /radical /approxequal 159,160c159,160 /.notdef /.notdef --- /Delta /lozenge I have no idea if it can have some negative effect if the same encoding name (TeXBase1Encoding) is loaded with different meanings in one (say Postscript) document. Thomas
Re: 2.99.6 beta release - problem with Irix Tar
On Fri, Dec 10, 2004 at 11:04:41PM -0500, ezra peisach wrote: tar: @LongLink - cannot create -- Permission denied Oops... I think that I have accidently unpacked a context distribution (cont-tmf.zip) while I was inside texk/dvipng. The files there are not used for anything, so even if some of them are broken due to the 100 chars limit, that should not hurt. If you want to repair your source tree: cd texk/dvipng rm -rf context doc fonts metapost scripts tex web2c Sorry for that mess and thanks for your notice! Thomas
2.99.6 beta release
This 2.99.6 fixes all issues that came up with the 2.99.5 release: build problems with texinfo and dvipng (and dvipng should now even work on ULTRIX) and updates pdftex to 1.20b-rc5. As usual: please, send me feedback, even if it all just works :-) Files (to appear soon): ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-2.99.6.20041211-beta.tar.gz ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmf-2.99.6.20041211-beta.tar.gz ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-2.99.6.20041211-beta.tar.gz or ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-src-2.99.6.20041211-beta.tar.gz ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmf-2.99.6.20041211-beta.tar.gz ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-2.99.6.20041211-beta.tar.gz Changes in the src area: * dvipng update (1.3 - 1.4) * update pdftex to 1.20b-rc5 * fix typo (== instead of =) in texk/dvipng/configure.in * clean up --with-x-toolkit (split for mf / xdvi) * texinfo 4.7.91 (pretest version) Changes in the texmf area: * update texdoc.php * add context's .sty files for LaTeX support Happy testing + have fun! Thomas
Re: 2.99.5 release
On Thu, Dec 09, 2004 at 03:38:15PM +0100, Erik Frisk wrote: ./configure: test: unknown operator == That is that line if test $ac_cv_header_inttypes_h == yes; then which should be if test $ac_cv_header_inttypes_h = yes; then Thanks for reporting it! Thomas
Re: 2.99.5 release
On Thu, Dec 09, 2004 at 12:29:52PM +0100, Martin Schröder wrote: No etex anymore? Shouldn't pdfetex take the role? Same goes for pdftex. I guess that some old fmtutil.cnf shadows the new one in $TEXMFMAIN. After removing it, just run fmtutil --all texlinks and you should have the missing symlinks and formats. Thomas
Re: ConTeXt configuration directories
On Thu, Dec 09, 2004 at 11:44:25AM +0100, Frank Küster wrote:
is it really necessary that ConTeXt has two config directories,
It has two directories which are named config. That's a fact. I don't
know if users have to change the files in there. I guess that this does
not really happen.
$ diff -u tetex-texmf-2.99.5.20041207-beta/context/config/texexec.{ini,rme}
Hans distributes texexec.rme as some kind of template for the real
texexec.ini file which defines the settings as appropriate for the TeX
system. For some reason, Hans' defaults perfectly match teTeX, so I
usually just copy texexec.rme to texexec.ini.
Thomas
Re: 2.99.5 release
On Wed, Dec 08, 2004 at 12:37:53PM -0500, ezra peisach wrote: Compiles aok under Irix 6.5.20m - using native compilers and GNUmake. Ok, thanks. (system make bombs out in texk somewhere)... Hm... Can you please help with some more information? Simple tests: xdvi, latex, mpost, pdflatex appears to work. Good, thanks. Suggestion - in the INSTALL file - it says to type ls | wc -l - I get 27. Perhaps ls -1 | wc -l would be more appropriate to confirm the number of binaries. I was not aware that there are versions of ls where -1 is needed even when the output does not go directly to a terminal. Thanks. realtively fast (xdvi, dvips) until I set the paper type in PDFTEX. This forces a rebuild of the formats... Looks like most of pdftex no longer uses a pdftex.cfg file. Everything has to be set with primitives at the macro level. And, to be able to make global settings, these changes have to be dumped into the format files. Thomas
Re: Update ConTeXt 2004-12-06 for teTeX 2.99.4-beta: map files
On Tue, Dec 07, 2004 at 09:48:43PM +0100, VnPenguin wrote: Thank you so much for your very quick reply! For the 2.99.5 release (just released), I had to figure out what to do anyway :-) Thomas
Re: 2.99.4 release
a) there is a man page for lambda but no binary/tool gets installed texk/web2c/fmtutil.in has that line. It might be, however, that some older fmtutil.cnf file shadows the newly installed one. b) installation to $DESTDIR vs. $prefix : for make install the makefiles in utils/texinfo and utils/texi2html use both $DESTDIR and $prefix for destination location while most other tools/directories install to (only) $prefix. These use automake and follow the GNU standards and that's also where Olaf want's to go to (switch to autoconf) with web2c. teTeX-3.0 will use something like web2c-7.5.3 (with some smaller changes), so a uniform usage of DESTDIR / prefix for all packages in teTeX-3.0 is not feasible. Thomas
Re: 2.99.4.20041128 and Vietnamese Unicode problem
I don't know where is the problem ? tetex-beta (unicode support) or vntex ? The utf8 support of LaTeX's inputenc package is more restricted than the support provided by the unicode package. This is more a known limitation than a bug. Thomas
Re: 2.99.4 with non writeable local TeX dir
On Mon, Nov 29, 2004 at 02:26:30PM +0100, Willi Burmeister wrote: I tried 2.99.4 today and a got this error: The planned new scripts (texconfig, updmap, fmtutil) will only write to $TEXMFCONFIG (configuration stuff) resp. $VARTEXMF (generated runtime files such as formats, map files). The default for these will be the same as $TEXMFMAIN, but any user / administrator will be able to change this (texmf.cnf file or envvar). The benefits of this will be: - $TEXMFDIST, $TEXMFLOCAL and $HOMETEXMF will not be touched (unless explicitly requested, e.g. by setting TEXMFCONFIG=$HOMETEXMF) - non-admin users can use texconfig (after setting TEXMFCONFIG and VARTEXMF envvars), updmap and fmtutil The new texconfig will no longer generate any files on startup (e.g. missing format files) or generate copies of config files. Thomas
Re: 2.99.3 beta and texdoc on macosx
On Fri, Nov 26, 2004 at 11:38:12PM +0100, Sven de Vries wrote:
1) it checks whether TEXDOCVIEW_dvi/ps is set the by the user and then
using that,
That's already the case, since we are using these ${var=value} constructs
which do not change $var if the variable has been set before.
BTW, the script already has a comment which explains this:
# Viewing programs, according to filename extension. (You can
# override or add to them by setting environment variables).
# MacOS X: does not have X11 by default, so give dvi a low priority
Otherwise:
2) checking whether xdvi/gv is in the defaultpath
(and perhaps whether X11 is running)
otherwise
3) using open
Something like this?
if test x$DISPLAY != x type -p xdvi /dev/null 21; then
: ${TEXDOCVIEW_dvi='xdvi %s '}
else
: ${TEXDOCVIEW_dvi='open %s'}
fi
if test x$DISPLAY != x type -p gv /dev/null 21; then
: ${TEXDOCVIEW_ps='gv %s '}
else
: ${TEXDOCVIEW_ps='open %s'}
fi
Thomas
Re: [2.99.3] ./configure[597]: syntax error at line 74 : `{' unmatched
Configure fails on Irix 6.5:
...
./configure[597]: syntax error at line 74 : `{' unmatched
Oops... I just found it in my PROBLEMS file:
- SGI Irix 6.5.22m, ./configure fails with:
...
loading cache .././config.cache
./configure[588]: syntax error at line 73 : `{' unmatched
Work-around: install fw_bash and set CONFIG_SHELL=/usr/freeware/bin/bash
(environment variable) for the configure run.
It might in fact be a problem with /bin/sh (being broken) on some versions
of Irix...
Thomas
Re: texhash in updmap
In fact, I can't imagine the situation that we don't have output files. The destination directory is not fixed. For example, if you copy the updmap.cfg file to $HOMETEXMF/web2c, the script will also use $HOMETEXMF as the destination directory for the generated map files. I plan to change the strategy that texconfig / fmtutil / updmap use for writing their config files and variable runtime data and th new strategy will make it even easier to switch... What I can offer is to check the existence of the output files on startup and not to run texhash in the end. So, for the usual case of It sounds good. Will you do me a favor? Yes. Thomas
Re: 2.99.3 beta and dvipdft
Is the following better?
s=$s${s+ }'`echo x$1 | sed [EMAIL PROTECTED]@@; s@'@'\''@g`'
this fixes the problem on the sun.
Next try (script attatched).
Thomas
#!/bin/sh
#
# dvipdft, completely rewtitten by Thomas Esser for teTeX after an
# original script by Mark A. Wicks
#
# Copyright (C) 2004 Thomas Esser. Public Domain.
#
# This script makes a first (fast) run with dvipdfm, then calls gs on
# the resulting pdf file to get the thumbnails and finally calls dvipdfm.
test -f /bin/sh5 test -z $RUNNING_SH5 \
{ UNAMES=`uname -s`; test x$UNAMES = xULTRIX; } 2/dev/null \
{ RUNNING_SH5=true; export RUNNING_SH5; exec /bin/sh5 $0 ${1+$@}; }
unset RUNNING_SH5
test -f /bin/bsh test -z $RUNNING_BSH \
{ UNAMES=`uname -s`; test x$UNAMES = xAIX; } 2/dev/null \
{ RUNNING_BSH=true; export RUNNING_BSH; exec /bin/bsh $0 ${1+$@}; }
unset RUNNING_BSH
# hack around a bug in zsh:
test -n ${ZSH_VERSION+set} alias -g '${1+$@}'='$@'
progname=dvipdft
version=1084580267 # seconds since `00:00:00 1970-01-01 UTC'
# date '+%s' (with GNU date)
help=Usage: $progname [options] dvifile
$progname --help
$progname --version
tmpdir=${TMP-/tmp}/$progname.$$
quoteit()
{
echo x$1 | sed [EMAIL PROTECTED]@@; s@'@'''@; [EMAIL PROTECTED]@'@;
[EMAIL PROTECTED]@'@
}
# process args 1 to (n-1) and scan for a -o argument keep a properly
# quoted backup of the options, so that they can be restored later
unset s outfile
while test $# -gt 1; do
case $1 in
-o) outfile=$2;;
esac
s=$s${s+ }`quoteit $1`
shift
done
echo $s
# check for last argument
case $1 in
--help)
echo $help
exit 0
;;
--version)
echo $progname version $version by Thomas Esser
exit 0
;;
)
echo $help
exit 1
;;
esac
# args 1 to (n-1) are gone, last one must be the dvifile
dvifile=$1
# outfile comes from -o argument, else it is derived from the input file
test -n $outfile \
|| outfile=`echo $dvifile | sed '[EMAIL PROTECTED]/@@; [EMAIL PROTECTED]@@;
[EMAIL PROTECTED]@.pdf@'`
# outBase is part of the thumbname files just as dvipdfm expects them:
outBase=`echo $outfile | sed '[EMAIL PROTECTED]@@'`
# restore args 1 to (n-1):
eval set x $s; shift
# Run dvipdfm with the fastest options for the first pass
echo $progname: running dvipdfm $@ -e -z0 $dvifile 2
dvipdfm $@ -e -z0 $dvifile || {
echo $progname: dvipdfm failed. 2
exit 1
}
# before we create the tmpdir, set trap for cleanup
trap '
rm -rf $tmpdir
exit 1
' 1 2 3 7 13 15
mkdir $tmpdir || {
echo $progname: failed to create temp directory. 2
exit 1
}
# if outBase contains a /, we might need to create a directory
case $outBase in
*/*)
fq=`echo $tmpdir/$outBase | sed 's@//@/@g; s@/[^/]*$@@'`
test -d $fq || mkdir -p $fq || {
echo $progname: failed to create temp thumbnail directory. 2
rm -rf $tmpdir
exit 1
}
;;
esac
# run gs
echo $progname: running gs 2
gs -r10 -dNOPAUSE -dBATCH -sDEVICE=png256 \
-sOutputFile=$tmpdir/$outBase.%d $outfile || {
echo $progname: gs failed. 2
rm -rf $tmpdir
exit 1
}
# run dvipdfm with the users specified options for the last pass
echo $progname: running dvipdfm -dt $@ $dvifile 2
TMP=$tmpdir dvipdfm -dt $@ $dvifile || {
echo $progname: dvipdfm failed. 2
rm -rf $tmpdir
exit 1
}
# nothing failed, so cleanup and report success to caller
rm -rf $tmpdir
exit 0
Re: Mathtime Plus problem
Is this intentional?
Yes.
I fixed the problem by running latex on mathtime.ins and installing
the *.fd files.
That's exactly right.
teTeX comes with a free replacement font for mathtime (belleek), but there
is no free replacement for the mathtime plus fonts. The .fd files that
are distributed with teTeX match the font set that is included with teTeX.
I have generated the .fd files by running latex mathtime.ins, too,
but I have made use of the following feature of the .ins file:
% If you have the `basic' MathTime font set, and *not*
% the extended MathTime Plus (or MathTime Complete) set
% then you should uncomment this line so that LaTeX does
% not attempt to access bold math fonts, or the new MTSYN
% font.
%
%\def\mtplus{}
Thomas
Re: 2.99.3 beta and dvipdft
On Tue, Nov 16, 2004 at 10:21:09AM +0100, Sven de Vries wrote:
seems like dvipdft is broken:
Works for me on SuSE Linux 9.1.
l43 of that script looks odd. has anybody encountered this too?
Is the following better?
s=$s${s+ }'`echo x$1 | sed [EMAIL PROTECTED]@@; s@'@'\''@g`'
Thomas
Re: texhash in updmap
(a) touch output files I have a problem with that. If something goes wrong, empty files might remain. What I can offer is to check the existence of the output files on startup and not to run texhash in the end. So, for the usual case of a non-initial run of updmap, the user puts his flies into place, then runs texhash und updmap. Since updmap detects that it does not add files to any texmf tree, it won't call texhash again. Thomas
Re: 2.99.2 xdvik does not detect multiple definition handling on IRIX
I'll try to release a fixed version of xdvik this evening that Thomas can put into the next tarball - sorry for the hassle ... Anything new besides the twi fixes to aclocal.m4 and that one fix to xdvi.h? These, I already have now... Thanks, Thomas
Re: 2.99.2 release
On Sun, Nov 07, 2004 at 11:32:26AM -0500, ezra peisach wrote: This is fixed in newer versions of the gd library... But here is the fix... (essentially remove inline). Thanks for mentioning the new version. I think that I'll just switch to that. I see several portability fixes and security fixes (overflow checking). Thomas
Re: 2.99.2 release
On Sat, Nov 06, 2004 at 11:19:55PM +, Robin Fairbairns wrote: pulled to cambridge and pushed out again... Amazingly fast again... You're the hero of this night! :-) Thomas
2.99.2: xdvik fix
A fix for xdvik. Thomas diff -ru tetex-src-2.99.2.20041106-beta/texk/xdvik/xdvi.h tetex-src/texk/xdvik/xdvi.h --- tetex-src-2.99.2.20041106-beta/texk/xdvik/xdvi.h2004-11-06 23:57:28.0 +0100 +++ tetex-src/texk/xdvik/xdvi.h 2004-11-07 01:31:03.427914881 +0100 @@ -992,9 +992,9 @@ * See the comment in events.c for further explanations. */ #if LD_ALLOWS_MULTIPLE_DEFINITIONS -# define XDVI_XT_XDVI_XT_TIMER_HACK 1 +# define XDVI_XT_TIMER_HACK 1 #else -# define XDVI_XT_XDVI_XT_TIMER_HACK 0 +# define XDVI_XT_TIMER_HACK 0 #endif /* globally used GUI stuff */
Re: 2.99.1: new beta / almost release candidate
Hi Reinhard,
/usr/lib/gcc-lib/i386-slackware-linux/egcs-2.91.66/include/stdbool.h:9: parse error
before `0'
Please, retry with the path that is attatched.
WARNING: You must install texinfo.tex and txi-??.tex manually,
Ok, I'll ask Karl for a switch to turn it off or something else...
fmtutil: attempting to create localized format using pool=tex-pl and tcx=il2-pl.
I use some tool to generate the localized pool files and this tool no
longer works with the changes format of the pool files. So, for now, you
have to disable the *-pl formats in fmtutil.cnf.
make: *** [install] Error 1
Have a look at the top-level Makefile. It runs texlinks and updmap
after fmtutil. Since your installation aborted, you have to run them by
hand now.
Thanks for the report!
Thomas
--- tetex-src-beta-2.99.1.20041026/texk/dvipng/color.c 2004-07-01 14:29:17.0
+0200
+++ tetex-src/texk/dvipng/color.c 2004-10-30 11:28:40.0 +0200
@@ -24,6 +24,9 @@
/
#include dvipng.h
+#if HAVE_ALLOCA_H
+# include alloca.h
+#endif
/*
* Color. We delete and recreate the gdImage for each new page. This
--- tetex-src-beta-2.99.1.20041026/texk/dvipng/dvipng.h 2004-10-23 21:45:08.0
+0200
+++ tetex-src/texk/dvipng/dvipng.h 2004-11-03 19:30:10.0 +0100
@@ -53,6 +53,10 @@
typedef unsigned long long uint64_t;
#endif
+#ifdef HAVE_STDBOOL_H
+#include stdbool.h
+#endif
+
#ifndef INT32_MIN
#define INT32_MIN (-2147483647-1)
#endif
@@ -63,7 +67,18 @@
#include gd.h
#ifdef HAVE_LIBKPATHSEA
-#include kpathsea/kpathsea.h
+# include kpathsea/kpathsea.h
+/* boolean is an enum type from kpathsea/types.h loaded in
+ kpathsea/kpathsea.h, use it as fallback */
+# ifndef __bool_true_false_are_defined
+# define bool boolean
+# endif
+#else
+/* This is really insufficient, currently no path searches will work
+ without kpathsea anyway */
+typedef int bool;
+# define true (bool) 1
+# define false (bool) 0
#endif
#ifdef HAVE_FT2
@@ -77,18 +92,6 @@
#include t1lib.h
#endif
-#ifdef HAVE_STDBOOL_H
-#include stdbool.h
-#else
-typedef int bool;
-#ifndef true
-#define true (bool) 1
-#endif
-#ifndef false
-#define false (bool) 0
-#endif
-#endif
-
#ifndef HAVE_VPRINTF
# ifdef HAVE_DOPRNT
# define vfprintf(stream, message, args) _doprnt(message, args, stream)
--- tetex-src-beta-2.99.1.20041026/texk/dvipng/enc.c2004-07-01 14:26:37.0
+0200
+++ tetex-src/texk/dvipng/enc.c 2004-10-30 11:31:14.0 +0200
@@ -24,6 +24,9 @@
/
#include dvipng.h
+#if HAVE_ALLOCA_H
+# include alloca.h
+#endif
struct encoding* encodingp=NULL;
--- tetex-src-beta-2.99.1.20041026/texk/dvipng/font.c 2004-07-01 14:25:43.0
+0200
+++ tetex-src/texk/dvipng/font.c2004-10-30 11:31:23.0 +0200
@@ -24,6 +24,9 @@
/
#include dvipng.h
+#if HAVE_ALLOCA_H
+# include alloca.h
+#endif
void CheckChecksum(uint32_t c1, uint32_t c2, const char* name)
{
@@ -188,7 +191,7 @@
}
#ifdef HAVE_FT2_OR_LIBT1
-inline char* kpse_find_t1_or_tt(char* filename)
+char* kpse_find_t1_or_tt(char* filename)
{
char* filepath = kpse_find_file(filename, kpse_type1_format, false);
#ifdef HAVE_FT2
--- tetex-src-beta-2.99.1.20041026/texk/dvipng/fontmap.c2004-07-01
14:24:46.0 +0200
+++ tetex-src/texk/dvipng/fontmap.c 2004-10-30 11:21:53.0 +0200
@@ -29,7 +29,7 @@
static struct filemmap psfont_mmap;
static struct psfontmap *psfontmap=NULL;
-inline char* newword(char** buffer, char* end)
+char* newword(char** buffer, char* end)
{
char *word,*pos=*buffer;
Re: [PATCH] Build on Solaris 8 with Sun's compilers
On Fri, Oct 29, 2004 at 05:12:09PM +0200, Jan-Åke Larsson wrote: Well, color.o had a missing alloca symbol that was fixed when I added the include to dvipng.h . I didn't track it down further, sorry. Silly me. Will be fixed in the next version. Don't forget enc.c and font.c. They use the TEMPSTR macro which uses alloca(), too. It might be possible toBah. I'll just remove the inline function-specifier. Ok, I have done the same in the copy included in teTeX. Thomas
Re: [PATCH] Build on Solaris 8 with Sun's compilers
On Thu, Oct 28, 2004 at 09:59:37AM -0700, Jason Riedy wrote: Appended is a patch that lets the current pretest (2.99.1.20041026) build, pass tests, and seem to work on Solaris 8 with Sun's 8.0 compilers. Thanks a lot for testing, and submitting this patch. I have fixed the places in my code and forwarded the changes further upspream (xpdf, pdftex, dvipng). Thomas
Re: 2.99.1: new beta / almost release candidate
Here's a small patch to fix the etex, pdftex and pdfetex and builds. Thanks. Applied and submitted upstream. noninteractively. I read you plan a rewrite, but wanted to let you know anyway. That kind of silly stuff is exactly what I want to get rid off. Thanks for your patch, Thomas
Re: 2.99.1: new beta / almost release candidate
Hi Jan, I'm quite happy with this release, it [cross] builds out of the box for Cygwin (except for the usual libtool handholding). Ah, finally! I am glad to hear this. I've added is the UNC file name patch, see below. Ok, forwarded to Olaf. A big thank you, also to Olaf. Most of the integration was in Olaf's code and he has done all the work in this area. Only a few places have been left over for me (to fix or to forward somewhere else). Thomas
Re: 2.99.1: new beta / almost release candidate
From xpdf-3.00-5.2.100mdk.src.rpm, I see as
xpdf-3.00-CAN-2004-0888-0889.patch.bz2 (from Oct. 19). Does this work or
does it suffer from gcc optimizations? At least, one place (see below)
explicitly says that this works...
--- xpdf-3.00/xpdf/XRef.cc.CAN-2004-0888-0889 2004-10-19 08:58:47.0 -0600
+++ xpdf-3.00/xpdf/XRef.cc 2004-10-19 08:58:47.0 -0600
@@ -110,6 +110,10 @@
goto err1;
}
+ if (nObjects*sizeof(int)/sizeof(int) != nObjects) {
+error(-1, Invalid 'nObjects');
+goto err1;
+ }
objs = new Object[nObjects];
objNums = (int *)gmalloc(nObjects * sizeof(int));
offsets = (int *)gmalloc(nObjects * sizeof(int));
@@ -385,6 +389,10 @@
for (newSize = size ? 2 * size : 1024;
first + n newSize newSize 0;
newSize = 1) ;
+ if (newSize*sizeof(XRefEntry)/sizeof(XRefEntry) != newSize) {
+ error(-1, Invalid 'obj' parameters');
+ return gFalse;
+ }
if (newSize 0) {
goto err1;
}
@@ -493,6 +501,10 @@
goto err1;
}
if (newSize size) {
+if (newSize * sizeof(XRefEntry)/sizeof(XRefEntry) != newSize) {
+ error(-1, Invalid 'size' parameter.);
+ return gFalse;
+}
entries = (XRefEntry *)grealloc(entries, newSize * sizeof(XRefEntry));
for (i = size; i newSize; ++i) {
entries[i].offset = 0x;
@@ -580,6 +592,10 @@
for (newSize = size ? 2 * size : 1024;
first + n newSize newSize 0;
newSize = 1) ;
+if (newSize*sizeof(XRefEntry)/sizeof(XRefEntry) != newSize) {
+ error(-1, Invalid 'size' inside xref table.);
+ return gFalse;
+}
if (newSize 0) {
return gFalse;
}
@@ -714,6 +730,10 @@
if (!strncmp(p, obj, 3)) {
if (num = size) {
newSize = (num + 1 + 255) ~255;
+ if (newSize*sizeof(XRefEntry)/sizeof(XRefEntry) != newSize) {
+ error(-1, Invalid 'obj' parameters.);
+ return gFalse;
+ }
if (newSize 0) {
error(-1, Bad object number);
return gFalse;
@@ -741,6 +761,10 @@
} else if (!strncmp(p, endstream, 9)) {
if (streamEndsLen == streamEndsSize) {
streamEndsSize += 64;
+if (streamEndsSize*sizeof(int)/sizeof(int) != streamEndsSize) {
+ error(-1, Invalid 'endstream' parameter.);
+ return gFalse;
+}
streamEnds = (Guint *)grealloc(streamEnds,
streamEndsSize * sizeof(int));
}
--- xpdf-3.00/xpdf/Catalog.cc.CAN-2004-0888-08892004-01-21 18:26:45.0
-0700
+++ xpdf-3.00/xpdf/Catalog.cc 2004-10-19 09:02:12.0 -0600
@@ -64,6 +64,15 @@
}
pagesSize = numPages0 = (int)obj.getNum();
obj.free();
+ // The gcc doesnt optimize this away, so this check is ok,
+ // even if it looks like a pagesSize != pagesSize check
+ if (pagesSize*sizeof(Page *)/sizeof(Page *) != pagesSize ||
+ pagesSize*sizeof(Ref)/sizeof(Ref) != pagesSize) {
+error(-1, Invalid 'pagesSize');
+ok = gFalse;
+return;
+ }
+
pages = (Page **)gmalloc(pagesSize * sizeof(Page *));
pageRefs = (Ref *)gmalloc(pagesSize * sizeof(Ref));
for (i = 0; i pagesSize; ++i) {
@@ -191,6 +200,11 @@
}
if (start = pagesSize) {
pagesSize += 32;
+if (pagesSize*sizeof(Page *)/sizeof(Page *) != pagesSize ||
+ pagesSize*sizeof(Ref)/sizeof(Ref) != pagesSize) {
+ error(-1, Invalid 'pagesSize' parameter.);
+ goto err3;
+}
pages = (Page **)grealloc(pages, pagesSize * sizeof(Page *));
pageRefs = (Ref *)grealloc(pageRefs, pagesSize * sizeof(Ref));
for (j = pagesSize - 32; j pagesSize; ++j) {
Re: 2.99.1: new beta / almost release candidate
And I sent a patch for teTeX 2.0.2 which has not yet reached me again
through the list. I fear I might have sent a wrong patch that still
contained code that would be optimised away by compilers. Here's the
correct one.
Ok, I have looked at the mandrake patch and they protect memory
(re)allocations in Catalog.cc and XRef.cc. I now have applied a
fix similar to their one, but using comparisons like pagesSize =
INT_MAX/sizeof(Page*) instead of these funny constructs that some
compilers might optimize away.
Thanks for your help, Frank!
Thomas
diff -ru tetex-src-beta-2.99.1.20041026/libs/xpdf/xpdf/Catalog.cc
tetex-src/libs/xpdf/xpdf/Catalog.cc
--- tetex-src-beta-2.99.1.20041026/libs/xpdf/xpdf/Catalog.cc2004-01-22
02:26:45.0 +0100
+++ tetex-src/libs/xpdf/xpdf/Catalog.cc 2004-10-27 19:14:44.350094017 +0200
@@ -13,6 +13,7 @@
#endif
#include stddef.h
+#include limits.h
#include gmem.h
#include Object.h
#include XRef.h
@@ -64,6 +65,13 @@
}
pagesSize = numPages0 = (int)obj.getNum();
obj.free();
+ if ((pagesSize INT_MAX / sizeof(Page*)) ||
+ (pagesSize INT_MAX / sizeof(Ref)) ||
+ (pagesSize 0)) {
+error(-1, Invalid 'pagesSize');
+ok = gFalse;
+return;
+ }
pages = (Page **)gmalloc(pagesSize * sizeof(Page *));
pageRefs = (Ref *)gmalloc(pagesSize * sizeof(Ref));
for (i = 0; i pagesSize; ++i) {
@@ -191,6 +199,12 @@
}
if (start = pagesSize) {
pagesSize += 32;
+if ((pagesSize INT_MAX / sizeof(Page*)) ||
+(pagesSize INT_MAX / sizeof(Ref)) ||
+(pagesSize 0)) {
+ error(-1, Invalid 'pagesSize' parameter);
+ goto err3;
+}
pages = (Page **)grealloc(pages, pagesSize * sizeof(Page *));
pageRefs = (Ref *)grealloc(pageRefs, pagesSize * sizeof(Ref));
for (j = pagesSize - 32; j pagesSize; ++j) {
diff -ru tetex-src-beta-2.99.1.20041026/libs/xpdf/xpdf/XRef.cc
tetex-src/libs/xpdf/xpdf/XRef.cc
--- tetex-src-beta-2.99.1.20041026/libs/xpdf/xpdf/XRef.cc 2004-01-22
02:26:45.0 +0100
+++ tetex-src/libs/xpdf/xpdf/XRef.cc2004-10-27 19:16:04.407374178 +0200
@@ -16,6 +16,7 @@
#include stddef.h
#include string.h
#include ctype.h
+#include limits.h
#include gmem.h
#include Object.h
#include Stream.h
@@ -107,6 +108,11 @@
first = obj1.getInt();
obj1.free();
+ if ((nObjects = INT_MAX / sizeof(int)) ||
+ (nObjects 0)) {
+error(-1, Invalid 'nObjects');
+goto err1;
+ }
objs = new Object[nObjects];
objNums = (int *)gmalloc(nObjects * sizeof(int));
offsets = (int *)gmalloc(nObjects * sizeof(int));
@@ -373,6 +379,11 @@
for (newSize = size ? 2 * size : 1024;
first + n newSize;
newSize = 1) ;
+ if ((newSize = INT_MAX / sizeof(XRefEntry)) ||
+ (newSize 0)) {
+error(-1, Invalid 'obj' parameters');
+return gFalse;
+ }
entries = (XRefEntry *)grealloc(entries, newSize * sizeof(XRefEntry));
for (i = size; i newSize; ++i) {
entries[i].offset = 0x;
@@ -475,6 +486,11 @@
newSize = obj.getInt();
obj.free();
if (newSize size) {
+if ((newSize = INT_MAX / sizeof(XRefEntry)) ||
+(newSize 0)) {
+ error(-1, Invalid 'size' parameter);
+ goto err0;
+}
entries = (XRefEntry *)grealloc(entries, newSize * sizeof(XRefEntry));
for (i = size; i newSize; ++i) {
entries[i].offset = 0x;
@@ -555,6 +571,11 @@
for (newSize = size ? 2 * size : 1024;
first + n newSize;
newSize = 1) ;
+if ((newSize = INT_MAX / sizeof(XRefEntry)) ||
+(newSize 0)) {
+ error(-1, Invalid 'newSize');
+ return gFalse;
+}
entries = (XRefEntry *)grealloc(entries, newSize * sizeof(XRefEntry));
for (i = size; i newSize; ++i) {
entries[i].offset = 0x;
@@ -683,6 +704,11 @@
if (!strncmp(p, obj, 3)) {
if (num = size) {
newSize = (num + 1 + 255) ~255;
+if ((newSize = INT_MAX / sizeof(XRefEntry)) ||
+(newSize 0)) {
+ error(-1, Invalid 'obj' parameters.);
+ return gFalse;
+}
entries = (XRefEntry *)
grealloc(entries, newSize * sizeof(XRefEntry));
for (i = size; i newSize; ++i) {
@@ -705,6 +731,11 @@
} else if (!strncmp(p, endstream, 9)) {
if (streamEndsLen == streamEndsSize) {
streamEndsSize += 64;
+if ((streamEndsSize = INT_MAX / sizeof(int)) ||
+(streamEndsSize 0)) {
+ error(-1, Invalid 'endstream' parameter);
+ return gFalse;
+}
streamEnds = (Guint *)grealloc(streamEnds,
streamEndsSize * sizeof(int));
}
2.99.0 beta
Well, after some time of silence, we now have a major update. TeX Live
is frozen now and time has come to get things released...
The following parts are the major missing points of my todo list. I
think that some of them won't make it into 3.0 and will come later.
- update TETEXDOC (more release notes etc.)
- doc/index.php etc.
- update teTeX-FAQ (changes location of map / enc files)
- update ncurses
- dvipng-1.2
- add texconfig paper for setting a4 or letter for all supported
applications
- texconfig rewrite (initialization, place to store config data etc.)
- add afm2pl
My plan is still to get 3.0 out this year. There will be a release
candidate before the final version.
In this beta, lots of things have changed again. Large packages have been
added to the texmf tree (such as beamer) and important things have changed
in the src area (Peter has fixed an important bug in etex / pdfetex).
updmap has been improved, too. I have even started to update TETEXDOC...
Files (to appear soon):
ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.99.0.20041023.tar.gz
ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.99.0.20041023.tar.gz
ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-beta-2.99.0.20041023.tar.gz
or
ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.99.0.20041023.tar.gz
ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.99.0.20041023.tar.gz
ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-beta-2.99.0.20041023.tar.gz
Changes in the src area:
* started to update TETEXDOC (but not yet finished)
* dvips bugfix (debian bug #277499 about mis-parsing of font
map files)
* t1lib.ac: use LIBS not LDFLAGS
* added TEXINPUTS.utf8mex to texmf.in
* update texdoctk (v.0.6.0)
* xdvik-22.84.4 (fixes a few bugs)
* add a few missing manpages (contributed by Frank Kuester /
Claire Connelly)
* set locale according to LC_CTYPE in texk/makeindexk/genind.c
for correct tolower() behaviour
* sync with TeX Live, lots of significant changes:
- pdfTeX 1.20a
- eTeX version 2.2 (hopefully final)
- Aleph 3.141592-1.15-2.1-rc4
* dvips version is now 5.94b (no code change)
* dvitomp.ch: a more complete implementation of the color stack.
* tex.ch: Revise the file:line:error code to produce the correct
filename in the presence of \scantokens (an e-TeX extension)
* different format of banner strings (web2c now comes last)
* updates in documentation (e.g. web2c.texi and manpages)
* mktexfmt.8 - fmtutil.8 (fixed broken link to fmtutil.1)
* add lamed (latex based on aleph), change lambda (latex based
on omega) in fmtutil.in
* update texinfo.tex
* sync with TeX Live; mostly changefile rearrangements
* rm dvipsk/tex/dvipsmac.tex dvipsk/tex/dvips.tex
* call updmap on make install (if updmap.cfg file is installed)
* pdftex 1.20a-rc7.2
Changes in the texmf area:
* footbib added back (now LPPL)
* amslatex update; more docs now pdf instead of dvi
* replace doc/latex/latex2e-html by chunced html version of
tex-refs (see http://www.miwie.org/tex-refs/)
* update acronym, beamer, eepic, exam, chngpage.sty, koma-script
threeparttable.sty, pict2e, showlabels, Catalogue, ps4pdf, leftidx
* update context to version 2004.10.19, additional bugfix for
core-job.tex from Hans
* update ltnews.cls (better using \pdfoutput now)
* updated doc/pdftex with files from TeX Live; pdfTeX-FAQ removed
(too many obsolete parts in it)
* update listings, psnfss
* update pstricks-add
* update minitoc
* update fancyhdr
* update microtype, textcase
* add files for utf8mex
* letterspacing.tex (from Phil Taylor) removed for license reasons
* removed tex/lambda/base/ltpatch.ltx
* update doc/index.{php,html}
* we now have the context manual: cont-eni.pdf (not exactly sure
when it was added)
* update pstricks
* update package microtype
* package xkeyval added
* add package microtype (LaTeX interface to micro-typographic
extensions of pdfTeX)
* update psnfss to version 9.2 and added FPL fonts
* update eplain
* update context to version 2004.8.18
* added slashed.sty by Carlisle
* update package listings (to version 1.3)
* added iso-8859-7.def
* added package filecontents
* update texinfo.tex
* rm tex/plain/dvips/dvipsmac.tex
* update pstricks
* update pict2e, dkhyphen
* get rid of symlinks in texmf tree:
- remove updmap
Re: Some really unimportant bits
On Mon, Sep 27, 2004 at 10:26:54AM +0200, Frank Küster wrote: odvicopy (Omega related) odvips (Omega related) odvitype (Omega related) In Debian, somebody decided to create symlinks to the respective manpages without the o. Does that make sense? Yes, better than no manpage at all. I just have made the change in my sources. Thanks, Thomas
Re: Some really unimportant bits
On Mon, May 03, 2004 at 01:16:55PM +0200, Frank Küster wrote: - A couple of binaries don't have manpages - or rather the links to I have now added some manpages from Frank and Claire (thanks!). The remaining missing ones are now: a2ping (perl script; converts graphics to pdf) aleph (eTeX enhanced omega) ctangle (Tangle WEBFILE with CHANGEFILE into a C/C++ program) cweave (Weave WEBFILE with CHANGEFILE into a TeX document) fdf2tan (perl script; belongs to ConTeXt) mptopdf (perl script; belongs to ConTeXt) lamed (symlink to aleph for latex format) odvicopy (Omega related) odvips (Omega related) odvitype (Omega related) omfonts (Omega related) otangle (Omega related) pdfcrop (perl script for cropping pdf graphics/files) xdvi-motif.bin (real name of xdvi binary) Thomas
Re: Some really unimportant bits
On Mon, May 03, 2004 at 09:10:22PM +0200, Frank Küster wrote: some weeks ago there was a discussion on the TDS mailing list about a change in TDS. In the future, platform independent programs (Perl, Java, Ruby-scripts) will go into $TEXMF/scripts. The scripts will be called by wrappers which are in the bin/[platform] directory. Then you don't have to patch texdoctk any more, the wrapper can check whether Tk.pm exists. This is a bit slower but much easier to maintain. Will this be in tetex-3.0? tetex-3.0 will allow to store scripts in texmf trees (kpathsea supports this), but this does not mean that all scripts will go into some texmf tree. I'll move scripts that can better be maintained inside some texmf tree, such as thumbpdf which is released together with thumbpdf.tex on CTAN. Having the script together with the tex file eliminates an unnecessary dependency between texmf and src tarball. As for texdoctk, I have not yet decided... Thomas
Re: Some really unimportant bits
On Sun, Sep 26, 2004 at 05:09:27PM +0200, Thomas Esser wrote: ctangle (Tangle WEBFILE with CHANGEFILE into a C/C++ program) cweave (Weave WEBFILE with CHANGEFILE into a TeX document) These are covered by cweb.1, so they can be marked as done as well. Thomas
Re: Cygwin-cross patch#3 for 2.96.4-1
FWIW, I never understood why tetex copies sources of external util and lib packages, in every distribution I have seen they are all disabled anyway (and packaged directly from upstream). I really wonder if teTeX contains its own *mandatory* libraries. One aim of teTeX it to have something easy to install for the *end user* which is a different situation from a distribution maker. If there are bugs, I'd be happy to hear about them to get them fixed. enough people are using this to justify the duplication of efforts? Which duplication of efforts? I don't reprogram the libraries, I just include them. Not much work. Alltogether certainly less work that the cumulated work that end users have if they need to find out hot to get / build / install these libraries on their own. Thomas
2.96.7 pretest
Yet anther src-update... There have been some important pdftex / etex
related updates and the changefiles have been rearranged to increase
the mainability. Fortunately, TeX 3.141592 related changes get properly
merged into etex / pdfetex.
I have had a quick look at Jan's Cygwin-Patches, but not yet adopted
anything. But, I will do this, i.e. either fix things if it affects my
stuff or forward the relevant patches upstream.
Files (to appear soon):
ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.7.20040721.tar.gz
or
ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.7.20040721.tar.gz
Changes in the src area:
* changes from Olaf Weber / Peter Breitenlohner / Han The Thanh /
Martin Schröder
- Updating etex-2.1 to TeX 3.141592 (not yet finished). This shall eventually
become etex-2.2.
- Updating pdftex to TeX 3.141592 (glue and leader box rounding plus some
typos). In doing so I have removed a few modules that were identical copies
of those for TeX; without these copies the leader box rounding problem
wouldn't have existed.
- Adapting pdfetex (and pdfxtex) to etex changes.
- Rewrite of build rules for {etex,pdf*tex}.{web,ch} and removal of obsolete
files from the dependencies. Removal of some spurious dependencies for
pdf*tex.web.
- pdftex 1.20a-rc7.1
* create symlink for mktexfmt.$(manext)
* fixed broken dvipng configure.in
* pdfcrop added to texk/tetex/Makefile.in
Happy testing + have fun!
Thomas
Re: 2.96.6 release (small update)
On Wed, Jul 21, 2004 at 11:12:35AM +0200, Jan Nieuwenhuizen wrote: dvipng's author complained about debug being off by default whereas his setting is on by default. Strange. IMHO, that's for the packager or distribution to decide and worry about, not for the upstream author. My wording was not accurate. It was more like a comment rather than a complaint. Anyway, I do agree to add the debug option. I would much rather have upstream worry about configure and compile and portability issues: include a configure check for libgen.h and my prototype patch, Jan-Åke? Do you listen? Did you look at Jan's patch? and upgraded to autoconf2.50. That autoconf-2.13 is what I currently need inside texk (for things like sinclude), so the old version is the fault of the old setup we (me, Olaf Weber, ...) currently use. This will improve sometime (maybe in a year from now). Actually, the original dvipng distribution uses autoconf 2.59. Thomas
2.96.6 release (small update)
This small update (only src tarball this time; use with texmf tarball from last release) fixes a few problems. Files (to appear soon): ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.6.20040715.tar.gz or ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.6.20040715.tar.gz Changes in the src area: * fixes to xdvik (22.84.3), etex, enctex * web2c: changes related to bug adresses in various programs * fix duplicate -o args when linking Happy testing + have fun! Thomas
2.95.5 on CTAN
Thanks to Robin's quick help, the files are now available on CTAN as well. The (tetex) master site is still offline, so we don't have md5sum files at the moment. Here are the links: ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.5.20040711.tar.gz ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.96.5.20040711.tar.gz ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-beta-2.96.5.20040711.tar.gz and ftp://dante.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.5.20040711.tar.gz ftp://dante.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.96.5.20040711.tar.gz ftp://dante.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-beta-2.96.5.20040711.tar.gz and ftp://tug.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.5.20040711.tar.gz ftp://tug.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.96.5.20040711.tar.gz ftp://tug.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-beta-2.96.5.20040711.tar.gz Thanks, Robin! Thomas
Re: teTeX 2.96.4 TIE
Thomas Esser writes: On Thu, Jul 08, 2004 at 08:16:05PM +0200, Hartmut Henkel wrote: E. g. where is $TIE normally set? texk/web2c/Makefile.in: TIE = @TIE@ texk/web2c/configure.in KPSE_CROSS_PATH_PROG(TIE, tie, ./tie) How come it wasn't set? (Dunno.) In fact i can't find any of these lines in the tetex-src-beta-2.96.4.20040627.tar.gz files. But I believe it _has_ worked... Oops... My reply was based on my current version of the code which is ahead a little (by changes from Olaf regarding cross compilation support). In 2.96.4, I see direct invocations of ./tie in pdfetexdir/pdfetex.mk and etexdir/etex.mk. These fragments end up in web2c/Makefile. So, I don't think that $(TIE) is set anywhere. But, this will change with the next version... Thomas
Re: teTeX 2.96.4 TIE
On Thu, Jul 08, 2004 at 08:16:05PM +0200, Hartmut Henkel wrote: E. g. where is $TIE normally set? texk/web2c/Makefile.in: TIE = @TIE@ texk/web2c/configure.in KPSE_CROSS_PATH_PROG(TIE, tie, ./tie) Thomas
Re: hyphenation patterns in language.dat: Why not enable all?
On Mon, Jul 05, 2004 at 09:04:34AM +0200, Frank Küster wrote: czech czhyph.tex % depends on csplain being installed This just means that czhyph.tex alone is not enough. czhyph.tex itself loads a few other files e.g. tex/csplain/czhyphen.tex. It does not need more than this. To make things more clear, I just have changed this into: czech czhyph.tex % depends on other files from tex/csplain Thomas
Re: mptopdf Problem
On Wed, Jun 30, 2004 at 09:28:35AM +0200, Hans Hagen wrote: pdfetex --help gives: While this may be true for the binaries that you have, this is not true for the current status of binaries that we currently compile for TeX Live and teTeX. So, for us, this means that some things in context just don't work. longer work with old binaries (because i use old and new ones, and because we cannot expect everyone to update, leats of all sysadms) But the new systems that we put up should work. anyhow, the fmt business will become messy anyway since either or not subpaths will be used (if not, there will be clashes) On teTeX, you can rely on fmtutil to do the job for you. You don't need to know where format files are stored and what extension they have. One thing if not supported and this is using one format with more than one engine. Thomas
Re: Package wishes: m-bibTeX and amsl/nath for ConTeXt
http://tex.aanhet.net/bib/ *Please*, convice the author to put his package to CTAN. I can't monitor a hundred different places... No, I am not going to get it from there. - amsl and nath for ConTeXt ftp://ftp.tex.ac.uk./tex-archive/macros/context/contrib/maths/ In ChangeLog: Thu May 27 19:22:52 CEST 2004 * update t-amsl.tex, t-nath.tex ... Fri Jan 16 20:27:54 CET 2004 * update context to version 2003.12.12, add t-amsl.tex and t-nath.tex Thomas
Re: xdvi.cfg
On Wed, May 12, 2004 at 09:54:53AM +0100, Stefan Ulrich wrote: Well, this refers to the `standalone' xdvik distribution - should we change the wording for teTeX? What I wanted to say is that users should more or less ignore the file in the source distribution. Well, I did not change Makefile.in (compared against the `standalone' xdvik distribution), so I don't think that the documentation needs to be different. I suggest to document what the Makefile does: it checks if there is no file or an original existing file (xdvi.cfg) and installs the supplied version in these cases, but not in case of an existing modified file. XDvi is always installed. current teTeX ... maybe we should simply remove the entire README.t1fonts (and pointers to it) from teTeX since it's too confusing? I think that some useful information is contained there and suggest to keep it. Thomas
Re: xdvi.cfg
On Wed, May 12, 2004 at 04:11:50PM +0400, Michail Vidiassov wrote: % for xdvik too. Users of version 1.0 of teTeX should additionally uncomment % the line with `psfonts.map' (preserving the order of the lines). % dvipsmap psfonts.map dvipsmap ps2pk.map Is the line with `psfonts.map' uncommented before installation? If you install from the current sources, you are not a user of teTeX *1.0*. Really, that version number matters here. In 1.0, ps2pk.map did not have all entries. It has just redefined the 35 Laserwriter fonts, so that they are no longer considered as driver build in. So, to answer your question: no it is not uncommented before installation. Still, this xdvi.cfg file is the right one to use now. Hm... Maybe, Stefan could make this: ... This file is not suitable for teTeX users with a version before 1.0. For using this file with teTeX-1.0.x, please uncomment the line with `psfonts.map' (preserving the order of the lines). For any later version of teTeX, no modification is needed. ... Thomas
Re: 2.96.3 beta -- `{' unmatched from configure on Irix
loading cache .././config.cache
./configure[588]: syntax error at line 73 : `{' unmatched
George, please, send me the files
config.cache
config.log
config.status
from the texk directory.
Has anybody else ever seen this? Has anybody else tried this or a
similar platform?
Thanks,
Thomas
new beta: 2.96.2
Well, some time has passed since the last beta and lots of things have changed again. The highlights: - updated prerelease of web2c-7.5.3 - stable pdfeTeX, Version 3.141592-1.20b-2.1 - stable xdvik-22.82 - dvipng added (has required to add gd library as well) - texinfo 4.7 - new newhelpindex.html by Joao Palhoto Matos I'd be happy to get any feedback (not only about problems/bugs), especially since lots of web2c internaly have changed (e.g. regarding 8bit clean \write) and a lot of new code was added (gd library and dvipng). So, if you can compile+run this successfully, please let me know some details about your platform (hardware, os, compiler etc.). Files (to appear soon): ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.2.20040510.tar.gz ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.96.2.20040510.tar.gz ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-beta-2.96.2.20040510.tar.gz or ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.2.20040510.tar.gz ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.96.2.20040510.tar.gz ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-beta-2.96.2.20040510.tar.gz Changes in the src area: * pdfeTeX, Version 3.141592-1.20b-2.1 * top-level Makefile: lo longer strip any binary * A huge change, to bring the teTeX source in sync with Olaf's working sources for web2c. These changes implement: - ini* and vir* links are no longer created - xord/xchr/xprn are always dumped - 8bits option - pool_name and engine_name parametrization of mf and mp - halt-on-error option - setting $engine in the environment - all formats have .fmt extension - quoted filenames so filenames with spaces can work * larger INBUFSIZE in afm2tfm.c * updates in fmtutil for Olaf's changes * minor fix in fmtutil * fix bug in usage string of makeindex * add dvipng and the gd library * update to xdvik-22.82 * major update: merge web2c-7.5.3 (prerelease) from Olaf; includes switch to new tex.web * bugfix in updmap manpage * update texinfo.tex * sync dvipsk with TeX Live * added README.texdoctk (old version, needs updating) * update to texinfo 4.7 * bugfix for pdftex scripts not being installed * pslatex improved * add mode dpdfezzz (8000dpi) to mktexpk for -Ppdf * make aleph the default engine for lambda (the LaTeX team has made a request for e-TeX based engines) * texk/tetex/texdoctk.man fixes (by Eric S. Raymond) * cygwin related fix in texconfig (for dialog) * add manpage for kpsewhere (thanks, Frank!) * clean up texk/tetex/Makefile for manpage symlinks (initex, virtex etc. are no longer installed, so we don't need manpages for this). * fix location for language.def in texconfig * texconfig: texmf-dist split requires different has_latex implementation * texconfig: now really use TEXMFLOCAL and not LOCALTEXMF. * get rid of MakeTeXPK (mktexpk is still there) Changes in the texmf area: * slightly improve comments in dvips' config.ps * newhelpindex.html by Joao Palhoto Matos * add package preview (from preview-latex) and ps4pdf * add enctex tables + documenation (from ftp://math.feld.cvut.cz/pub/olsak/enctex/enctex.tar.gz) * update context (to version 2004.4.9); fixes a very important bug (latex/3657) and now adops to TDS rearrangements (for texmf/scripts) * add definitions for a0, a1 and a2 to dvips' config.ps * add type1 version of doublestroke fonts * update texinfo.tex, footmisc, Catalogue, wasychr.mf, frhyph.tex * update chapterbib.sty, fontinstallationguide.pdf, l2tabuen, stmaryrd, ccaption, hyphenat, layouts, supertabular * updates in babel (dutch+frenchb) * update mfpic, footmisc * update fontname map files * add alias for ec.enc * update caption2 * rm xdvi/XDvi (installed by make install) * update texdoctk database * add Estonian hyphenation patterns (eehyph.tex), license: lppl * update lm fonts (ver. 0.92) * update url.sty, uk-tex-faq, lshort * update doc/pdftex/base/pdftex-syntax.txt * update modes.mf * remove elatex.ini * rewrite the code in .ini files to work with non-pdftex based engines * add contnav.map * add natural.tcx * update texinfo.tex Happy testing + fave fun! Thomas
Re: tetex 2.96 beta texlinks uses nonstandard access
Irix does not provide the program access Most other system don't as well. access is not supposed to come from the os, but from kpathsea. However, the program has been renamed to kpseaccess. Since I always have an access binary somewhere around, I did not notice. So, the real fix is to replace access - kpseaccess in texlinks. Thanks for reporting, Thomas
Re: [2.96.1] problem: latex generates pdf rather than dvi
which makes me wonder if the hyperref package somehow overrode the choice
of dvi output.
Running latex on a minimal document that loads hyperref, e.g.
\documentclass{minimal}
\usepackage{hyperref}
\begin{document}
Hi
\end{document}
shows that hyperref can happily work with our current setup.
So, I guess that it is some other package or the document itself that
sets \pdfoutput=1.
Thomas
Re: [2.96.1] problem: latex generates pdf rather than dvi
\ifx\pdfoutput\undefined
\usepackage[dvips]{graphicx}
% \usepackage[dvips,a4paper,breaklinks]{hyperref}
\else
\usepackage[pdftex]{graphicx}
\usepackage[pdftex,a4paper,pageanchor,hyperfigures,breaklinks,%
colorlinks,citecolor=blue,urlcolor=red,linkcolor=magenta]{hyperref
}
\pdfcompresslevel=9
\fi
Make this
\usepackage{ifpdf}
\ifpdf
\usepackage[dvips]{graphicx}
\else
\usepackage[pdftex]{graphicx}
...
\fi
IMHO, it is wrong to assume that the user wants pdf output if the pdftex
engine is detected. It is in the documentation of pdftex for *ages*
(forever I think) that pdftex can well be used to produce dvi output.
Thomas
Re: new beta release: 2.96.0.20040321
On Mon, Mar 22, 2004 at 10:08:11AM +0100, Tobias Burnus wrote: Is there any chance that Aleph (formally, eOmega) gets included? (I would prefer to have it in teTeX rather than only on the TeXLive CD.) Well, I just have added it (RC2), so: yes! :-) Thomas
Re: new beta release: 2.96.0.20040321
The list of binaries in INSTALL seems to be wrong. I only get 121: Yes, that's still wrong. I have to update it... Thanks, Thomas
new beta: 2.96.1.20040322
Well, several people have had build problems with yesterday's beta, so this one follows rather quickly. The most significant change (besides xdvik fixes) is the addition of Aleph (=Omega-1.15 merged with e-TeX). Files (to appear soon): ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.1.20040322.tar.gz ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.96.1.20040322.tar.gz or ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.1.20040322.tar.gz ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.96.1.20040322.tar.gz Changes in the src area: * xdvik fixes (upgrade to 22.81.2-CVS3) * add Aleph * texconfig stores modified config files in $TEXMFLOCAL instead of $TEXMFMAIN resp. $VARTEXMF * sh reautoconf to get the default prefix right Changes in the texmf area: * add aleph.ini * downgrade pdftex.def (by Martin Schröder's request) * update ConTeXt to version 2004.3.19 This is a beta release, but I really need feedback on the changes and rearrangements. Happy testing + have fun! Thomas
new beta release: 2.96.0.20040321
On the way to teTeX-3.0, I make this interim beta. Several important things have been updated and we have again important changes in implementation details. The most significant changes are: - pdfetex is the main engine - LaTeX2e 2003/12/01 - pdftex 1.20a-rc1 - xdvik 22.81.1 Most important todo: - wait for final pdftex 1.20a - update the documentation - update updmap (it should help users with the new localtion for map files) The new main tex engine means that etex and pdftex are no longer build. pdfetex is used where tex, etex and pdftex have been used before (except that tex is still good old original tex, but e.g. latex is now pdfetex). One other important change is that pdftex no longer reads a special config file at startup. Everything is done through the normal tex input machanism, so all *.ini files have been modified to read a new pdftexconfig.tex which replaces pdftex.cfg. Files (to appear soon): ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.0.20040321.tar.gz ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.96.0.20040321.tar.gz ftp://cam.ctan.org/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-beta-2.96.0.20040321.tar.gz or ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-src-beta-2.96.0.20040321.tar.gz ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmf-beta-2.96.0.20040321.tar.gz ftp://ftp.dante.de/tex-archive/systems/unix/teTeX-beta/tetex-texmfsrc-beta-2.96.0.20040321.tar.gz Changes in the src area (slightly beatified from Changelog): * adjust texconfig for pdftex.cf - pdftexconfig.tex * ps2pk manpage was unaware of kpathsea; fixed by Frank Küster * update pdftex to version 1.20a-rc1 * upgrade to xdvik-22.81.1-RC2 * patch from Akira Kakuto (and approved by Olaf): take file_line_error_style from texmf.cnf * dvipsk/dvips.h: larger STRINGSIZE (just always use the large value that was previously only used for odvips) * updfate texinfo.tex * pdftex merged from TeX Live, changes at some places to make pdfetex the main TeX engine; no longer symlinks for ini*tex and vir*tex * kpathsea: TEXMFSCRIPTS and LIGFONTS added * rm -rf utils/texinfo/djgpp/CVS * add angle(0,0) to the list of non-fatal errors in mktexpk Changes in the texmf area (slightly beatified from Changelog): * remove pdftex.cfg * icelandic hyphenation added (icehyph.tex, LPPL) * update c-cmr.mf from CTAN:fonts/cm/sauter/base/c-cmr.mf * update psgo * enc files move to dvips subdirectory * rename pdfconfig.tex to pdftexconfig.tex and adjust .ini files * update uk-tex-faq * update Catalogue, fontname files (map / enc), l2tabu, changebar, fancyhdr, listings, ntgclass * added pict2e * latex update: base toos graphics psnfss cyrillic babel * update texinfo.tex * add pdfconfig.tex (replaces pdftex.cfg) * change all *.ini files (except tex.ini) to read pdfconfig.tex * rm -f doc/fonts/marvosym/marvosym.pfb * update texnames.sty, ithyph.tex * rm ./fonts/map/dvips/base/psfonts.map (file exists in fonts/map/dvips/updmap) * mv fonts/type1/public/cs/README doc/fonts/cs * update texinfo.tex, hyperref * add import.sty * adjust A4size definitions in dvips's config.ps (see IEEE testflow document for explanation) and change pdftexDownloadBase14 to true in updmap.cfg. * update texinfo.tex, hyperref, subfig * add tfm files of doublestroke fonts * mv fonts/ocp/public/oinuit/ omega/ocp * update custom-bib (only README.413 + merlin.mbs) * add x.mf / x.mp (to exit, just like x.tex) This is a beta release, but I really need feedback on the changes and rearrangements. Happy testing + fave fun! Thomas
Re: new beta release: 2.96.0.20040321
[note: Cc: changed from tetex to tetex-pretest] PREFIX [/usr/local] Is this an intended change? /usr/local/teTeX used to be the default prefix. No, that came from playing with various things. I had uncommented AC_PREFIX_DEFAULT(/usr/local/teTeX) if test x$prefix = xNONE; then prefix=$ac_default_prefix ac_configure_args=$ac_configure_args --prefix=$ac_default_prefix fi in tetex.ac, and finally decided to keep it. But, obviously, I have forgotten to recreate the configure file. The easiest fix is to specify --prefix=/usr/local/teTeX explicitly this time. For the next time, this will be the default again. I hope that nobody messes his /usr/local up... Sorry, Thomas
Re: texconfig should not stop when unimportant files are missing
Of course. But this is a general proposal - there are many reasons why a file is missing or moved, and it would be nice if texconfig would be more clever in which files it really needs. I know this is hard to I agree, although I don't give it a high priority... Thomas
