![](https://secure.gravatar.com/avatar/75dc6ef5c344c5b25200e44d4088ef2a.jpg?s=120&d=mm&r=g)
Hi, have you working installation of latest context and luatex on linux/amd64? Drop some info please. I have no problem with installation on 32bit machine. The same system (gentoo) on 64bit doesn't generate "cont-en.fmt". Lua and luatex are installed from source. Zdenek
![](https://secure.gravatar.com/avatar/cedc4d83943006afe3994fb0a0455f82.jpg?s=120&d=mm&r=g)
On Tue, 4 Sep 2007, Taco Hoekwater wrote:
Soprry, not yet (I have to reformat a machine)
In the meantime, you could just replace the executable with the i586 version: http://foundry.supelec.fr/frs/download.php/513/luatex-beta-0.10.2-linux-i586... Cheers, Peter -- http://pmrb.free.fr/contact/
![](https://secure.gravatar.com/avatar/cedc4d83943006afe3994fb0a0455f82.jpg?s=120&d=mm&r=g)
On Wed, 5 Sep 2007, zs wrote:
In the meantime, you could just replace the executable with the i586
I've tried precompiled binaries. They either do not run at all or behave identically.
You mean, format generation does not work with i586 binary neither? Strange, then it seems that you encounter another problem than the "luatex Segmentation fault" mentioned 3 weeks ago here on this list... No problem here (also linux-amd64) with default installation from http://pmrb.free.fr/texlive/ and replacement of luatex binary afterwards. Cheers, Peter -- http://pmrb.free.fr/contact/
![](https://secure.gravatar.com/avatar/a0f6b595b0e3b7f3abba19eeb8261fc2.jpg?s=120&d=mm&r=g)
have you working installation of latest context and luatex on linux/amd64? Drop some info please.
contextgarden.net runs on AMD64. It always has the latest version of ConTeXt, and a recent luatex (beta-0.11.0 as I speak). Everything is compiled from source.
I have no problem with installation on 32bit machine. The same system (gentoo) on 64bit doesn't generate "cont-en.fmt".
The garden uses Debian. And file tells me the luatex binary is actually 32-bit, although I wonder how this can be possible. What is your error message exactly? Arthur
![](https://secure.gravatar.com/avatar/75dc6ef5c344c5b25200e44d4088ef2a.jpg?s=120&d=mm&r=g)
The garden uses Debian. And file tells me the luatex binary is actually 32-bit, although I wonder how this can be possible.
Problem is on 64bit gentoo. On a 32bit (at home) it works.
What is your error message exactly?
Commands luatools --ini --compile --verbose cont-en and texmfstart texexec --make --all --luatex do not produce error message at all. But texmfstart texexec --luatex something.tex ends with: This is luaTeX, Version 3.141592-beta-0.11.0-2007082309 (Web2C 7.5.6) luatex: /tmp/luatex-cache/context/2bee42bddafb2afd5c3c1eaea23baaa5/formats/cont-en.fmt: No such file or directory LuaTools | running command: luatex --fmt="/tmp/luatex-cache/context/2bee42bddafb2afd5c3c1eaea23baaa5/formats/cont-en" --lua="/tmp/luatex-cache/context/2bee42bddafb2afd5c3c1eaea23baaa5/formats/cont-en.luc" "something.tex"
![](https://secure.gravatar.com/avatar/a0f6b595b0e3b7f3abba19eeb8261fc2.jpg?s=120&d=mm&r=g)
texmfstart texexec --make --all --luatex
do not produce error message at all.
This does not indicate that all the format files have been compiled. Are there where luatools looks for them? (/tmp/luatex-cache/context/2bee42bddafb2afd5c3c1eaea23baaa5/formats/cont-en.fmt) If not, can you send the exact output of texexec --luatex --make --all (maybe by private mail)? If yes, this is extremely weird. Arthur
![](https://secure.gravatar.com/avatar/565476157eb3895a9c928ffce26d988d.jpg?s=120&d=mm&r=g)
Arthur Reutenauer wrote:
texmfstart texexec --make --all --luatex
do not produce error message at all.
This does not indicate that all the format files have been compiled. Are there where luatools looks for them? (/tmp/luatex-cache/context/2bee42bddafb2afd5c3c1eaea23baaa5/formats/cont-en.fmt) If not, can you send the exact output of texexec --luatex --make --all (maybe by private mail)? If yes, this is extremely weird.
There was a bug in the binary that crashed the executable depending on the memory load location, so it is no sutprise that different machines gave different results. The current SVN trunk should have a fixed the problem. Best wishes, Taco
participants (4)
-
Arthur Reutenauer
-
Peter Münster
-
Taco Hoekwater
-
zs