Debian ConTeXt packages 2007.02.21
ConTeXt release 2007.02.21 is there ... On Son, 25 Feb 2007, Norbert Preining wrote:
deb http://www.tug.org/texlive/Debian/ context/ deb-src http://www.tug.org/texlive/Debian/ context/ BUT!!! You need the new lmodern 1.01 Debian package for this: deb http://www.tug.org/texlive/Debian/ lmodern/ deb-src http://www.tug.org/texlive/Debian/ lmodern/
I will upload both of these packages soon to Debian EXPERIMENTAL (!!!). Since Florent (main responsible for the lmodern package) and I agreed that we don't want to push lmodern 1.01 into etch (and there it is not necessary since there is no independent context package), the upload will go to experimental. As long as you have the above apt-get source nothing will change for you, but for people only using Debian/unstable source there will be no update.
If you have any questions please let me know!
Best wishes
Norbert
-------------------------------------------------------------------------------
Dr. Norbert Preining
Hi all!
I am currently uploading a whole new bunch of stuff to tug.org and
Debian/experimental:
For ConTeXt 2007.03.19-1 you have to go as usual to
deb http://www.tug.org/texlive/Debian/ context/
deb-src http://www.tug.org/texlive/Debian/ context/
but this new release depends on TeX Live 2007 for Debian, which can be
found here (and after some time (NEW processing) in
Debian/experimental):
deb http://www.tug.org/texlive/Debian/ tl2007/
deb-src http://www.tug.org/texlive/Debian/ tl2007/
there you also find tex-common 1.3 which is necessary for TeX Live 2007
packages. Finally you also need lmodern 1.01 which is available from
deb http://www.tug.org/texlive/Debian/ lmodern/
deb-src http://www.tug.org/texlive/Debian/ lmodern/
All these packages are uploaded to experimental, but some packages have
to go through the new queue, as they contain new binary packages
(texlive-xetex, -humanities, -science).
Please give it a twist, please report anything strange.
Thanks for testing!
Best wishes
Norbert
-------------------------------------------------------------------------------
Dr. Norbert Preining
Hi Norbert et al, I encountered a strange problem with metapost since my recent updates. I am using the apt-sources given in your post below and can't seem to make metapost happy, leaving it always complaining about wrong formats: ... This is MetaPost, Version 0.993 (Web2C 7.5.6) (/usr/share/texmf/web2c/natural.tcx) (Fatal mem file error; I'm stymied) ... I tried to purge and reinstall the whole enchilada from scratch, but to no avail. Afterwards I didn't do any special configurations, just # apt-get update # apt-get install context # apt-get install texlive-font-utils texlive-fonts-{recommended,extra} which gave me the following texlive packages: ||/ Name Version +++-=============================-========== ii texlive-base 2007-1 ii texlive-base-bin 2007-1 ii texlive-common 2007-1 ii texlive-doc-base 2007-1 ii texlive-font-utils 2007-1 ii texlive-fonts-extra 2007-1 ii texlive-fonts-recommended 2007-1 ii texlive-metapost 2007-1 and in /var/lib/texmf/web2c/metapost there are freshly generated metafun.mem and mpost.mem, no other mem-files on the system. A texexec --check results in ... TeXExec | TeXExec | version 6.2.0 - 1997-2006 - PRAGMA ADE/POD TeXExec | TeXUtil | version 9.1.0 - 1997-2005 - PRAGMA ADE/POD TeXExec | CtxTools | version 1.3.3 - 2004/2006 - PRAGMA ADE/POD TeXExec | TeXExec | testing interface en TeXExec | pdfTeX Version 3.141592-1.40.3 (Web2C 7.5.6) TeXExec | ConTeXt ver: 2007.03.19 11:20 MKII fmt: 2007.3.27 int: english/english Any hints hartelijk welcome. Jak. Norbert Preining wrote:
Hi all!
I am currently uploading a whole new bunch of stuff to tug.org and Debian/experimental:
For ConTeXt 2007.03.19-1 you have to go as usual to deb http://www.tug.org/texlive/Debian/ context/ deb-src http://www.tug.org/texlive/Debian/ context/ but this new release depends on TeX Live 2007 for Debian, which can be found here (and after some time (NEW processing) in Debian/experimental): deb http://www.tug.org/texlive/Debian/ tl2007/ deb-src http://www.tug.org/texlive/Debian/ tl2007/ there you also find tex-common 1.3 which is necessary for TeX Live 2007 packages. Finally you also need lmodern 1.01 which is available from deb http://www.tug.org/texlive/Debian/ lmodern/ deb-src http://www.tug.org/texlive/Debian/ lmodern/ All these packages are uploaded to experimental, but some packages have to go through the new queue, as they contain new binary packages (texlive-xetex, -humanities, -science).
Please give it a twist, please report anything strange.
Thanks for testing!
Best wishes
Norbert
------------------------------------------------------------------------------- Dr. Norbert Preining
Università di Siena Debian Developer Debian TeX Group gpg DSA: 0x09C5B094 fp: 14DF 2E6C 0307 BE6D AD76 A9C0 D2BF 4AA3 09C5 B094 ------------------------------------------------------------------------------- CORRIEMUCHLOCH (n.) Word describing the kind of person who can make a complete mess of a simple job like walking down a corridor. --- Douglas Adams, The Meaning of Liff _______________________________________________ dev-context mailing list dev-context@ntg.nl http://www.ntg.nl/mailman/listinfo/dev-context
plink wrote:
Hi Norbert et al,
I encountered a strange problem with metapost since my recent updates. I am using the apt-sources given in your post below and can't seem to make metapost happy, leaving it always complaining about wrong formats:
... This is MetaPost, Version 0.993 (Web2C 7.5.6) (/usr/share/texmf/web2c/natural.tcx) (Fatal mem file error; I'm stymied) ...
and in /var/lib/texmf/web2c/metapost there are freshly generated metafun.mem and mpost.mem, no other mem-files on the system.
Besides finding the wrong file, another possible explanation is that there is a mismatch between the array sizes in texmf.cnf that were used during mem dump time and the sizes that are to be used runtime. The 0.993 beta checks these values to make sure there is no change. This can happen for instance when you have a texmf.cnf like this: main_memory.mpost = 100000 main_memory.metafun = 2000000 In both cases, the -kpathsea-debug switch is useful. Either on the commandline, or in the environment. mpost -kpathsea-debug=6 or KPATHSEA_DEBUG=6 export KPATHSEA_DEBUG mpost The numeric value is a bitfield, the combination '6' should give you file operations and texmf.cnf value lookups Good luck, Taco
KPATHSEA_DEBUG=6 export KPATHSEA_DEBUG mpost
Or (in Bourne-shell syntax): KPATHSEA_DEBUG=6 mpost which sets the environment variable only for that invocation of mpost. -Sanjoy `Not all those who wander are lost.' (J.R.R. Tolkien)
Thanks for the hint, Taco! I found the following memory setting of the debian packages in debians configlet /etc/texmf/texmf.d/95NonPath.cnf: ... main_memory = 1000000 % words of inimemory available; also applies to inimf&mp extra_mem_top = 0 % extra high memory for chars, tokens, etc. extra_mem_bot = 0 % extra low memory for boxes, glue, breakpoints, etc. % ConTeXt is a memory hog... extra_mem_top.context = 2000000 extra_mem_bot.context = 4000000 main_memory.context = 1500000 main_memory.mpost = 1500000 ... Setting main_memory to the same size as main_memory.mpost fixes the problem (after an update-texmf and fmtutil-sys --refresh). However, now that I fixed it this way I do not understand why the same settings seem to work for other people, at least thats the way the Mpgraphic wiki page seems to point at. Sigh, one day when I grow up I'll be able to see into these dark TeX corners. Thanks very much anyhow! Taco Hoekwater wrote:
plink wrote:
I encountered a strange problem with metapost since my recent updates. I am using the apt-sources given in your post below and can't seem to make metapost happy, leaving it always complaining about wrong formats: ... This is MetaPost, Version 0.993 (Web2C 7.5.6) (/usr/share/texmf/web2c/natural.tcx) (Fatal mem file error; I'm stymied) ... and in /var/lib/texmf/web2c/metapost there are freshly generated metafun.mem and mpost.mem, no other mem-files on the system.
Besides finding the wrong file, another possible explanation is that there is a mismatch between the array sizes in texmf.cnf that were used during mem dump time and the sizes that are to be used runtime. The 0.993 beta checks these values to make sure there is no change.
This can happen for instance when you have a texmf.cnf like this:
main_memory.mpost = 100000 main_memory.metafun = 2000000
In both cases, the -kpathsea-debug switch is useful. Either on the commandline, or in the environment.
mpost -kpathsea-debug=6 or KPATHSEA_DEBUG=6 export KPATHSEA_DEBUG mpost
The numeric value is a bitfield, the combination '6' should give you file operations and texmf.cnf value lookups
Good luck,
Taco
Hi, plink wrote:
Thanks for the hint, Taco!
I found the following memory setting of the debian packages in debians configlet /etc/texmf/texmf.d/95NonPath.cnf:
... main_memory = 1000000 % words of inimemory available; also applies to inimf&mp extra_mem_top = 0 % extra high memory for chars, tokens, etc. extra_mem_bot = 0 % extra low memory for boxes, glue, breakpoints, etc.
% ConTeXt is a memory hog... extra_mem_top.context = 2000000 extra_mem_bot.context = 4000000 main_memory.context = 1500000 main_memory.mpost = 1500000 ...
Setting main_memory to the same size as main_memory.mpost fixes the problem (after an update-texmf and fmtutil-sys --refresh).
However, now that I fixed it this way I do not understand why the same settings seem to work for other people, at least thats the way the Mpgraphic wiki page seems to point at.
The change was made only in metapost 0.992, so until very recently, nobody was using it at all. The new version is now included in TeXLive 2007, but TL uses a different, corrected texmf.cnf itself. It looks like perhaps Norbert missed this change, and that that is why the texmf.cnf snippet in the debian packages is still wrong. (This is my fault, I should have told him) Norbert, can you fix this please? Best wishes, Taco
Taco Hoekwater wrote:
I encountered a strange problem with metapost since my recent updates. I am using the apt-sources given in your post below and can't seem to make metapost happy, leaving it always complaining about wrong formats: ... This is MetaPost, Version 0.993 (Web2C 7.5.6) (/usr/share/texmf/web2c/natural.tcx) (Fatal mem file error; I'm stymied) ... and in /var/lib/texmf/web2c/metapost there are freshly generated metafun.mem and mpost.mem, no other mem-files on the system. Besides finding the wrong file, another possible explanation is that there is a mismatch between the array sizes in texmf.cnf
plink wrote: that were used during mem dump time and the sizes that are to be used runtime. The 0.993 beta checks these values to make sure there is no change.
This can happen for instance when you have a texmf.cnf like this:
main_memory.mpost = 100000 main_memory.metafun = 2000000
In both cases, the -kpathsea-debug switch is useful. Either on the commandline, or in the environment.
mpost -kpathsea-debug=6 or KPATHSEA_DEBUG=6 export KPATHSEA_DEBUG mpost
The numeric value is a bitfield, the combination '6' should give you file operations and texmf.cnf value lookups
Good luck,
Taco
_______________________________________________ dev-context mailing list dev-context@ntg.nl http://www.ntg.nl/mailman/listinfo/dev-context
Hi Taco, hi all! On Don, 29 Mär 2007, Taco Hoekwater wrote:
main_memory = 1000000 % words of inimemory available; also applies to main_memory.context = 1500000 main_memory.mpost = 1500000
Norbert, can you fix this please?
Next version (1.4) of tex-common will have main_memory = 1500000.
But isn't this rubbish??? Why do we need
main_memory
and
main_memory.mpost
if both have to be the same??? What do I miss?
Best wishes
Norbert
-------------------------------------------------------------------------------
Dr. Norbert Preining
Norbert Preining wrote:
Hi Taco, hi all!
On Don, 29 Mär 2007, Taco Hoekwater wrote:
main_memory = 1000000 % words of inimemory available; also applies to main_memory.context = 1500000 main_memory.mpost = 1500000 Norbert, can you fix this please?
Next version (1.4) of tex-common will have main_memory = 1500000.
But isn't this rubbish??? Why do we need main_memory and main_memory.mpost if both have to be the same??? What do I miss?
No, we don't need both. In the older setups, there were multiple values for main_memory and main_memory.XXXX created, in an attempt to be conservative with memory usage for TeX when possible. But that size-difference did not work for metapost at all (thanks to either a different internal memory usage or because nobody ever looked into it and ust assumed that if it worked for TeX, it would work for metapost as well). Instead, it resulted in really strange and unpredictable bugs. So, in new metapost (>0.992) the value used during runtime has to be the same as the mem dumping value. It can still be main_memory.XXXX, but then you have to give a --progname switch during dump (and undump) so that both array sizes are always the same. Best, Taco
participants (4)
-
Norbert Preining
-
plink
-
Sanjoy Mahajan
-
Taco Hoekwater