Undefined control sequence. <argument> \c!gebied

Undefined control sequence. <argument> \c!gebied

[Thomas A.Schmitz]

OK, I tried several things to fix the error "Undefined control 
sequence. <argument> \c!gebied," but it persists. I wiped out my entire 
texmf.local tree and started from scratch because I thought the 
c!gebied was residing in some older file that ha dnot been overwritten. 
But even starting from scratch wouldn't amend the situation; after 
regenerating the format, I get the same error. However, grepping for 
"c!gebied" in the newly created tree gives only this result:

grep -r 'c!gebied' ./*
./tex/context/base/cont-sys.ori:% \setupexternalfigures 
[\c!gebied={e:/fig/eps,t:/mine/figs}]
./tex/context/user/cont-sys.rme:% \setupexternalfigures 
[\c!gebied={e:/fig/eps,t:/mine/figs}]

So this shouldn't be the problem since these two lines in cont-sys.rme 
are commented out. Does anybody have an idea where the error is coming 
from? There must be some old config file re-introducing it into the 
format file, but I can't seem to find it.

Baffled

Thomas

Re: Undefined control sequence. <argument> \c!gebied

[Lutz Haseloff]

Thomas A.Schmitz schrieb:
> OK, I tried several things to fix the error "Undefined control sequence. 
> <argument> \c!gebied," but it persists. I wiped out my entire 
> texmf.local tree and started from scratch because I thought the c!gebied 
> was residing in some older file that ha dnot been overwritten. But even 
> starting from scratch wouldn't amend the situation; after regenerating 
> the format, I get the same error. However, grepping for "c!gebied" in 
> the newly created tree gives only this result:
> 
> grep -r 'c!gebied' ./*
> ./tex/context/base/cont-sys.ori:% \setupexternalfigures 
> [\c!gebied={e:/fig/eps,t:/mine/figs}]
> ./tex/context/user/cont-sys.rme:% \setupexternalfigures 
> [\c!gebied={e:/fig/eps,t:/mine/figs}]

I had the same problem.
There where two texexec.pl, one in perltk (the old place)
and one in scripts/context/perl
In the new texexec.pl the command is \c!directory.
The Solution was to delete the perlscripts in the directory
perltk.

Hope this helps.

Lutz

> So this shouldn't be the problem since these two lines in cont-sys.rme 
> are commented out. Does anybody have an idea where the error is coming 
> from? There must be some old config file re-introducing it into the 
> format file, but I can't seem to find it.
> 
> Baffled
> 
> Thomas
> 
> _______________________________________________
> ntg-context mailing list
> ntg-context@ntg.nl
> http://www.ntg.nl/mailman/listinfo/ntg-context
> 

Re: Undefined control sequence. <argument> \c!gebied

[Thomas A.Schmitz]

Thanks for your reply. Unfortunately, that's not the problem here: 
since I deleted the entire tree, there is no second texexec.pl in 
web2c/context/perltk (which is where the scripts used to reside). In 
your case, the problem went away after deleting the old scripts? 
Interesting, and mysterious.

Best

Thomas
On Oct 11, 2004, at 2:55 PM, Lutz Haseloff wrote:

> I had the same problem.
> There where two texexec.pl, one in perltk (the old place)
> and one in scripts/context/perl
> In the new texexec.pl the command is \c!directory.
> The Solution was to delete the perlscripts in the directory
> perltk.
>
> Hope this helps.
>
> Lutz

Re: Undefined control sequence. <argument> \c!gebied

[Hans Hagen]

Thomas A.Schmitz wrote:

> OK, I tried several things to fix the error "Undefined control sequence. 
> <argument> \c!gebied," but it persists. I wiped out my entire 
> texmf.local tree and started from scratch because I thought the c!gebied 
> was residing in some older file that ha dnot been overwritten. But even 
> starting from scratch wouldn't amend the situation; after regenerating 
> the format, I get the same error. However, grepping for "c!gebied" in 
> the newly created tree gives only this result:
> 
> grep -r 'c!gebied' ./*
> ./tex/context/base/cont-sys.ori:% \setupexternalfigures 
> [\c!gebied={e:/fig/eps,t:/mine/figs}]
> ./tex/context/user/cont-sys.rme:% \setupexternalfigures 
> [\c!gebied={e:/fig/eps,t:/mine/figs}]
> 
> So this shouldn't be the problem since these two lines in cont-sys.rme 
> are commented out. Does anybody have an idea where the error is coming 
> from? There must be some old config file re-introducing it into the 
> format file, but I can't seem to find it.

strange, can you send me the log file?


Hans

-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
      tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
                                              | www.pragma-pod.nl
-----------------------------------------------------------------

Re: Undefined control sequence. <argument> \c!gebied

[Hans Hagen]

Lutz Haseloff wrote:

> I had the same problem.
> There where two texexec.pl, one in perltk (the old place)
> and one in scripts/context/perl
> In the new texexec.pl the command is \c!directory.
> The Solution was to delete the perlscripts in the directory
> perltk.

ah .. indeed, you need the texexec that matches the format!

this is one of the changes in the new tds: scripts are now under 
tex/texmf.../scripts;

one way to deal with such changes from now on, is to use texmfstart, i.e.:

texexec.bat ==

@echo off
texmfstart texexec.pl %*

texmfstart will always be adapted to the latest changes in tds etc and 
tries to be downward compatible



Hans

-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
      tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
                                              | www.pragma-pod.nl
-----------------------------------------------------------------

Re: Undefined control sequence. <argument> \c!gebied

[Thomas A.Schmitz]

Dear Hans and all,

again, thanks for your help! Everything is working as it should now. 
After asking on the TeX on OSX mailing-list, I found out that there was 
an experimental installation that already contains pdftex 1.20a, so I 
upgraded. I changed texmf.cnf as you suggested to Christopher Creutzig 
yesterday, regenerated mapfiles, and everything's bright again. But I 
was sweating and cursing... If it hadn't been for a complete backup of 
my old TeX installation on my second hard drive and your wonderful 
help, I would have had a lot of serious troubles. I guess that's a 
general problem with TeX installations: originally, they were made for 
UNIX mainframes maintained by knowledgeable sys admins who would likely 
know how to cope with such a situation. Nowadays, I assume that the 
majority of TeX users sit in front of their own box, and even though I 
can do some things on the command line and have by now acquired a 
reasonabel amount of TeXpertise, I would have been absolutely helpless 
without you and Gerben's great installer. So you're absolutely right: 
downward compatibility is broken; it is difficult to make local changes 
without wiping the entire installation. That's what you get from being 
impatient...

All best, and thanks

Thomas

On Oct 11, 2004, at 10:03 PM, Hans Hagen wrote:

> i'll try to keep texmfstart as compatibile as possible; there has been 
> a change in tds with regards to scripts (interestingly context was the 
> first to come with such scripts and the rule was to use /context/...); 
> when more package came with scripts tds was changed and now it's 
> /scripts/... ; the problem is that there is no strong urge to be 
> downward compatible (i tried to convince them in change but in vain) 
> and  since kpsewhich now uses another --format directive for such 
> scripts, downward compatibility in my opinion is broken (all scripts 
> that somehow use kpse to determine what to run will fail from now on, 
> unless changed); so, this is why i cooked up texmfstart: it tries to 
> be clever in determining what to call. (You cannot imagine how much of 
> my local and project trees and scripts i had to adapt; a lot of wasted 
> time)