metapost not invoked? mpgraph.1 not created

metapost not invoked? mpgraph.1 not created

[Patrick Gundlach]

Hi,

if I change the texexec perl script and add a exit 0 after line 
1767: 

[...]
#D   [       file=texexec.pl,
#D        version=2000.03.25,
[...]
              { print "                format : $MpFormat\n" ;
                $cmd .= " $MpPassString $MpFormatFlag$MpFormat " }
            $Problems = system ("$cmd $MpName" ) ;
	exit 0;  # <-----------------------
open (MPL,"$MpName.log") ;
while (<MPL>) # can be one big line unix under win
[...]

now I run 
texexec --interface=en --pdf mprun.tex
then
texexec --interface=en --mptex mpgraph.mp
and then
texexec --interface=en --pdf mprun.tex

the mpgraph.1 is generated.

only \runMPgraphicstrue is declared

this is the diff of the ls -l command between the first texexec 
and the second texexec command:

< -rw-rw-r--    1 gundlach gundlach      122 Feb 20 14:57 mpgraph.log
< -rw-rw-r--    1 gundlach gundlach        5 Feb 20 14:57 mpgraph.mp
---
> total 24
> -rw-rw-r--    1 gundlach gundlach      769 Feb 20 14:56 mpgraph.log
> -rw-rw-r--    1 gundlach gundlach     2221 Feb 20 14:56 mpgraph.mp
5d4
< -rw-rw-r--    1 gundlach gundlach      448 Feb 20 14:57 mpgraph.mpx
15,18d13
< -rw-rw-r--    1 gundlach gundlach     2221 Feb 20 14:56 tmpgraph.kep
< -rw-rw-r--    1 gundlach gundlach     2040 Feb 20 14:57 tmpgraph.log
< -rw-rw-r--    1 gundlach gundlach       82 Feb 20 14:57 tmpgraph.tmp
< -rw-rw-r--    1 gundlach gundlach      222 Feb 20 14:57 tmpgraph.tui

the < is after the second run.

...clueless :-)

Viele Gruesse,

  Patrick Gundlach

Re: metapost not invoked? mpgraph.1 not created

[Hans Hagen]

At 03:16 PM 2/20/01 +0100, Patrick Gundlach wrote:
>Hi,
>
>if I change the texexec perl script and add a exit 0 after line 
>1767: 
>
>[...]
>#D   [       file=texexec.pl,
>#D        version=2000.03.25,
>[...]
>              { print "                format : $MpFormat\n" ;
>                $cmd .= " $MpPassString $MpFormatFlag$MpFormat " }
>            $Problems = system ("$cmd $MpName" ) ;
>	exit 0;  # <-----------------------
>open (MPL,"$MpName.log") ;

I'm clueless too. Does anyone know what this means what this results in?
Some quit? Is this perl dependent? 

Also, can you check your MPXCOMMAND var in texmf.cnf? (set it to nothing) 

Hans
-------------------------------------------------------------------------
                                  Hans Hagen | PRAGMA ADE | pragma@wxs.nl
                      Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
 tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------

Re: metapost not invoked? mpgraph.1 not created

[Ed L Cashin]

Hans Hagen <pragma@wxs.nl> writes:

> At 03:16 PM 2/20/01 +0100, Patrick Gundlach wrote:
> >Hi,
> >
> >if I change the texexec perl script and add a exit 0 after line 
> >1767: 
> >
> >[...]
> >#D   [       file=texexec.pl,
> >#D        version=2000.03.25,
> >[...]
> >              { print "                format : $MpFormat\n" ;
> >                $cmd .= " $MpPassString $MpFormatFlag$MpFormat " }
> >            $Problems = system ("$cmd $MpName" ) ;
> >	exit 0;  # <-----------------------
> >open (MPL,"$MpName.log") ;
> 
> I'm clueless too. Does anyone know what this means what this results in?
> Some quit? Is this perl dependent? 

exit just terminates the perl program after running any "END" block.
The exit status of the program is the value of the expression
following "exit".  

So Patrick Gundlach has just told the perl program to terminate
immediately with an exit status of zero (success).

-- 
--Ed Cashin                    integrit file verification system
  ecashin@coe.uga.edu          http://integrit.sourceforge.net/

metapost not invoked? mpgraph.1 not created

[Patrick Gundlach]

Hi,

there is absolutely no hint about a misconfigurated PATH 
environment variable.

Viele Gruesse,

  Patrick Gundlach

metapost not invoked? mpgraph.1 not created

[Patrick Gundlach]

Hi,

> try to insert the absolute path to mpto here (for me it is 
> /opt/teTeX/bin/mpto)  If this works, there is a problem with your path.

OK, I really doubt that this is a 'path' problem. strace output gives 
me no hints of not found programs. Absolute path did not help.

> And please: try to "answer" the messages because not writing a reply 
> disturbs the threading ...

I *did* answer the message. Sorry your email reader is not 
capable of correct threading. (Kmail 2 bothers me with incorrect 
threading support, too.)

Viele Gruesse,

  Patrick Gundlach

metapost not invoked? mpgraph.1 not created

[Patrick Gundlach]

Hi,

I have tried to use the file with an older context format. There 
it worked :-):

This is pdfeTeX, Version 3.14159-13d-2.1 (Web2C 7.3.1)
entering extended mode
(tmpgraph.tex[/home/gundlach/TeX/texmf/pdftex/pdftex.cfg]

ConTeXt ver: 1999.5.11 fmt: 2001.2.1 int: english mes: english

texexec --interface=en --pdf mprun.tex
texexec --interface=en --mptex mpgraph.mp
only 
\runMPgraphicstrue
enabled

What about this extended mode (pdfetex)? Do I have to switch it
on? How do I switch it on?

Is it possible that it has to do with my context version? I have 
kept the texexec script, disabled my cont-en symbolic link to 
pdfetex, disabled pdfetex and removed my texmf/tex/context 
directory from my tex search path. I have also pointed the (e)fmt 
searchpath to the 'old' context.

Now I've tried  a new and old pdfetex with the old context format. The 
texexec --mptex mpgraph.mp gives an output.

ConTeXt ver: 1999.5.11 fmt: 2001.2.20 

texexec --verbose
             setting : 'MpToTeXExecutable' to 'mpto' for 'all'
               setting : 'MpToTeXExecutable' to 'mpto' for 'tetex'
          used setting : MpToTeXExecutable = mpto

gundlach@sarti:~/TeX/context/mprun$ which mpto
/usr/bin/mpto
gundlach@sarti:~/TeX/context/mprun$ mpto mpgraph.mp
\shipout\hbox{\smash{\hbox{\hbox{% line 3 mpgraph.mp
Hello world}\vrule width1sp}}}
\end{document}

Viele Gruesse,

  Patrick Gundlach

metapost not invoked? mpgraph.1 not created

[Patrick Gundlach]

Hi Hans and others,

> >mprun.log:systems        : please update your pdfTeX binaries

> eh, maybe, how old is your pdftex? On our site there are pointers to newer
> binaries (thanh has a linux one on his site). 
i've just compiled a new binary -> same problem.

This is pdfeTeX, Version 3.14159-13h-pretest-20010126-2.1 (Web2C 7.3.3.1)
I have tested another linux binary (precompiled) from tanh 
website (0.14f)

> I assume that you run texexec with --pdf ? What happens of you comment the
> \...true lines? 
'nothing' :-) no mpgraph.1 generated. When I only declare 
\runMPgraphicstrue and not \runMPTEXgraphicstrue metapost tells 
me:

--------------------------------------
Randomizer initialized to 780718373.This is MetaPost, Version 0.641 (Web2C 7.3.1)
(mpgraph.mp (/home/gundlach/TeX/texmf/metapost/context/mp-tool.mp)
(/home/gundlach/TeX/texmf/metapost/context/mp-spec.mp)
(/home/gundlach/TeX/texmf/metapost/context/mp-grph.mp)
(/home/gundlach/TeX/texmf/metapost/context/mp-page.mp)
(/home/gundlach/TeX/texmf/metapost/context/mp-core.mp)
>> mpgraph.mp
>> mpgraph.mpx
! Unable to make mpx file.
l.83 draw btex
               \framed [width=fit,align=middle]{\input tufte \relax } etex

[...]
-----------------------------------------

Just to make things clear:
this is a sample file that gives me headaches :-) (I have 
removed the \framed[....]{...} stuff to make sure that there is 
no problem here)

% also tried with % removed
%\runMPgraphicstrue
%\runMPTEXgraphicstrue
%\useMETAFUNformattrue

\starttext
\startuseMPgraphic{Text}
 draw btex Hello world etex ; 
\stopuseMPgraphic
\useMPgraphic{Text}
\stoptext

------------------------------------

I think that the line with btex--etex is problematic.
the mpgraph.log file has no error messages:
 This is MetaPost, Version 0.641 (Web2C 7.3.1) (mem=metafun 2001.2.19)  19 FEB 2001 13:00
**&metafun mpgraph
(mpgraph.mp )

There is just *no* output. The same(?) mpost run, with exactly the 
words '&metafun mpgraph' typed by hand gives me an errorfree 
output:

This is MetaPost, Version 0.641 (Web2C 7.3.1) (mem=metafun 2001.2.19)  19 FEB 2001 13:00
**&metafun mpgraph
(mpgraph.mp
{randomseed:=2508}
write0 = `mpgraph.mpo'.

 [1] )
1 output file written: mpgraph.1

When I try the test.tex mentioned in the metafun (beta) manual, 
the output file mpgraph.1 is generated.

\starttext 
\setupcolors[state=start] 
\startMPpage colorcircle(10cm,red,green,blue) ;
\stopMPpage \stoptext

-------------------------------
some more questions:

* is there a problem only having a cont-en.efmt (no pdftex 
version, only pdfetex)?

* mptopdf: There is a perl script  with this name 
which is in my $PATH  
(unix). There is also a mptopdf.efmt. Is this correct?

I have no idea where to look for more hints on this issue.

> Do others recognize this problem? 

Viele Gruesse,

  Patrick Gundlach

Re: metapost not invoked? mpgraph.1 not created

[Hans Hagen]

At 11:48 AM 2/20/01 +0100, Patrick Gundlach wrote:

>>> mpgraph.mp
>>> mpgraph.mpx
>! Unable to make mpx file.
>l.83 draw btex
>               \framed [width=fit,align=middle]{\input tufte \relax } etex
>
>[...]

It looks like texexec cannot locate / run "mpto"; can you run 

texexec --verbose 

and see if texexec reports the right mpto? 

The day is near that i kick out mpto and implement it myself in texexec.  

>There is just *no* output. The same(?) mpost run, with exactly the 
>words '&metafun mpgraph' typed by hand gives me an errorfree 
>output:

then you're lucky since that read has some nastly potential loopholes in it. 

>I have no idea where to look for more hints on this issue.

Don't despair, once it works, it keeps on working. It may be unix relatex
since it runs smoothly on all win machines that i spool tex onto. 

Are there unix people out there (i know there are) ? 

Hans
-------------------------------------------------------------------------
                                  Hans Hagen | PRAGMA ADE | pragma@wxs.nl
                      Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
 tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------

Re: metapost not invoked? mpgraph.1 not created

[Ed L Cashin]

Hans Hagen <pragma@wxs.nl> writes:

> Are there unix people out there (i know there are) ? 

The UN*X PATH variable works similarily to the DOS PATH variable, so
although I am a unix person out here, I don't have any special
insight.  (Of course, it's not necessary for mptopdf or the format
file to be literally in the PATH.)

-- 
--Ed Cashin                    integrit file verification system
  ecashin@coe.uga.edu          http://integrit.sourceforge.net/

Re: metapost not invoked? mpgraph.1 not created

[Marc van Dongen]

Ed L Cashin (ecashin@coe.uga.edu) wrote:

: Hans Hagen <pragma@wxs.nl> writes:
: 
: > Are there unix people out there (i know there are) ? 

Sorry Hans, I didn't spot your original query.

: The UN*X PATH variable works similarily to the DOS PATH variable, so

Not quite. In DOS it is by default made known to subprocesses.
In UNIX you have to explicitly say it should by uttering some
magic spell like:

PATH=/abra/ka/dabra:/abra/kebab/ra
eport PATH

or 

export PATH=/abra/ka/dabra:/abra/kebab/ra

if you're using bash or sh

or

setenv PATH /abra/ka/dabra:/abra/kebab/ra

if you're using (tc)sh

: although I am a unix person out here, I don't have any special
: insight.  (Of course, it's not necessary for mptopdf or the format
: file to be literally in the PATH.)

Even if

echo $PATH
sez
/abra/ka/dabra:/abra/kebab/ra

it may well be that because it is not exported its value may be
/usr/local/bin

Again, I missed Hans' original query, so I don't know if this
is relevant.

Regards,

Marc
-- 
     Marc van Dongen, CS Dept | phone:  +353 21 4903578
University College Cork, NUIC | Fax:    +353 21 4903113
  College Road, Cork, Ireland | Email: dongen@cs.ucc.ie

Re: metapost not invoked? mpgraph.1 not created

[Hans Hagen]

Hi, 

>Again, I missed Hans' original query, so I don't know if this
>is relevant.

It may be relevenat. I hope that the three of you can sort it out, so that
the solution can go into the faq. 

On our server i also found out that a bin can be in a path but when that
path has teh wrong flags it's skipped. Maybe the problem is in a
texmf-local tree with wrong access or not being in the path. 

Hans 
-------------------------------------------------------------------------
                                  Hans Hagen | PRAGMA ADE | pragma@wxs.nl
                      Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
 tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------

Re: metapost not invoked? mpgraph.1 not created

[Slawek Zak]

Hans Hagen <pragma@wxs.nl> writes:

> Hi, 
>  
> >Again, I missed Hans' original query, so I don't know if this
> >is relevant.
> 
> It may be relevenat. I hope that the three of you can sort it out, so that
> the solution can go into the faq. 

I don't know whether this is not the same problem I've had. When you
uncompress the context archive it doesn't have the +x bits which make
the scripts being executable.

In texexec 2.5 you have:

sub RunPerlScript
  { my ($ScriptName, $Options) = @_ ;
    if ($dosish)
      { if (-e "$TeXScriptsPath$ScriptName.pl")
          { system ("perl $TeXScriptsPath$ScriptName.pl $Options") } }
    else
      { system ("$ScriptName $Options") } }

Which makes perl try to run the script on UNIX systems without
checking if it's executable (texutil was -x in my case) and, which is
more problematic, doesn't check for possible execution errors. I've
proposed some time ago to change the call to something like:

  ...
  if (! -x $ScriptName) {
        die("Nonexecutable script, fix the permisions for $ScriptName\n");
  } else {
         system ("$ScriptName $Options") or die $!;
  }
  ...

What do you think?

/S

Re: metapost not invoked? mpgraph.1 not created

[Slawek Zak]

Slawek Zak <zaks@prioris.mini.pw.edu.pl> writes:

>   ...
>   if (! -x $ScriptName) {
>         die("Nonexecutable script, fix the permisions for $ScriptName\n");

Ouch! It should be full path to $ScriptName not just the tail.

/S

Re: metapost not invoked? mpgraph.1 not created

[Patrick Gundlach]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="us-ascii", Size: 336 bytes --]

Hi Slawek,

> I don't know whether this is not the same problem I've had. When you
> uncompress the context archive it doesn't have the +x bits which make
> the scripts being executable.

No, everything is executable. The problem still persists. But with 
\protectbufferstrue everything is ok.

-- 
Viele Grüße, 

    Patrick Gundlach

Re: metapost not invoked? mpgraph.1 not created

[Ed L Cashin]

Slawek Zak <zaks@prioris.mini.pw.edu.pl> writes:

...
> Which makes perl try to run the script on UNIX systems without
> checking if it's executable (texutil was -x in my case) and, which is
> more problematic, doesn't check for possible execution errors. I've
> proposed some time ago to change the call to something like:
> 
>   ...
>   if (! -x $ScriptName) {
>         die("Nonexecutable script, fix the permisions for $ScriptName\n");
>   } else {
>          system ("$ScriptName $Options") or die $!;
>   }
>   ...
> 
> What do you think?

I think it's not necessary to be that specific, since perl knows what
the last error was (it's in the $! variable), and you're limiting
yourself to just one error.  

If the $ScriptName wasn't there at all, for example, your error
message would be misleading.  Instead, let things take care of
themselves:

  $ cp /bin/ls .
  $ chmod 644 ls
  $ perl -e 'die "Error: could not run ls: $!" if system "./ls"'
  Error: could not run ls: Permission denied at -e line 1.

... and anyone who's familiar with UN*X has seen "Permission denied"
when the permission to execute is not set.

-- 
--Ed Cashin                   PGP public key:
  ecashin@terry.uga.edu       http://www.terry.uga.edu/~ecashin/pgp/

Re: metapost not invoked? mpgraph.1 not created

[Ed L Cashin]

Marc van Dongen <dongen@cs.ucc.ie> writes:

...
> it may well be that because it is not exported its value may be
> /usr/local/bin

Usually some system-wide resource configuration ("rc") file has
already exported the $PATH shell variable.  (It doesn't hurt to export
it again, though.)

On the other hand, it's very common for /usr/local/bin to *not* be in
the $PATH, so it might be handy to assign the PATH explicitly in your
shell's rc file.

-- 
--Ed Cashin                    integrit file verification system
  ecashin@coe.uga.edu          http://integrit.sourceforge.net/

Re: metapost not invoked? mpgraph.1 not created

[Marc van Dongen]

Ed L Cashin (ecashin@coe.uga.edu) wrote:

: > it may well be that because it is not exported its value may be
: > /usr/local/bin
: 
: Usually some system-wide resource configuration ("rc") file has
: already exported the $PATH shell variable.  (It doesn't hurt to export
: it again, though.)

The point is, if people add things to the path *after* it is
exported or add thing to the path as part of a shell-script,
these additions will not be made known to subshells. You
*have* to export it again. 

Regards,

Marc

Re: metapost not invoked? mpgraph.1 not created

[Ed L Cashin]

Marc van Dongen <dongen@cs.ucc.ie> writes:

> Ed L Cashin (ecashin@coe.uga.edu) wrote:
> 
> : > it may well be that because it is not exported its value may be
> : > /usr/local/bin
> : 
> : Usually some system-wide resource configuration ("rc") file has
> : already exported the $PATH shell variable.  (It doesn't hurt to export
> : it again, though.)
> 
> The point is, if people add things to the path *after* it is
> exported or add thing to the path as part of a shell-script,
> these additions will not be made known to subshells. You
> *have* to export it again. 

That's not true for Bourne-style shells:

    $ foo=hi
    $ perl -e 'print $ENV{"foo"} . "\n"'

    $ export foo
    $ perl -e 'print $ENV{"foo"} . "\n"'
    hi
    $ foo=hi,there
    $ perl -e 'print $ENV{"foo"} . "\n"'
    hi,there

I wouldn't know about C shells.

-- 
--Ed Cashin                    integrit file verification system
  ecashin@coe.uga.edu          http://integrit.sourceforge.net/

Re: metapost not invoked? mpgraph.1 not created

[Marc van Dongen]

Ed L Cashin (ecashin@coe.uga.edu) wrote:

: That's not true for Bourne-style shells:
: 
:     $ foo=hi
:     $ perl -e 'print $ENV{"foo"} . "\n"'
:     
:     $ export foo
:     $ perl -e 'print $ENV{"foo"} . "\n"'
:     hi
:     $ foo=hi,there
:     $ perl -e 'print $ENV{"foo"} . "\n"'
:     hi,there

It *is* for Bourne-style shells. If you run perl like
you do, foo is dereferenced in the current shell.

$ cat mytest
#!/bin/sh
echo ${MYVAR}
$ MYVAR=howsagoin
$ echo ${MYVAR}
howsagoin
$ mytest

$ export MYVAR
$ mytest
howsagoin
$

Regards,

Marc

Re: metapost not invoked? mpgraph.1 not created

[Marc van Dongen]

Marc van Dongen (dongen@cs.ucc.ie) wrote:

: Ed L Cashin (ecashin@coe.uga.edu) wrote:
: 
: : That's not true for Bourne-style shells:
: : 
: :     $ foo=hi
: :     $ perl -e 'print $ENV{"foo"} . "\n"'
: :     
: :     $ export foo
: :     $ perl -e 'print $ENV{"foo"} . "\n"'
: :     hi
: :     $ foo=hi,there
: :     $ perl -e 'print $ENV{"foo"} . "\n"'
: :     hi,there
: 
: It *is* for Bourne-style shells. If you run perl like
: you do, foo is dereferenced in the current shell.

Okay. I didn't properly study the first part of the example,
so forget about my comment.

Regards,

Marc

Regards,

Marc

Re: metapost not invoked? mpgraph.1 not created

[Uwe Koloska]

You wrote on Dienstag, 20. Februar 2001 11:48:
>Hi Hans and others,
>
>> >mprun.log:systems        : please update your pdfTeX binaries
>>
>> eh, maybe, how old is your pdftex? On our site there are pointers to
>> newer binaries (thanh has a linux one on his site).
>
>i've just compiled a new binary -> same problem.

How did you install your TeX system?  I suppose you use tetex.  Then if you 
install a new version of pdftex, you have to make sure that all paths are 
set correctly -- I don't know wether other versions use the same 
configuration than tetex ...

Normally by installing tetex, the location of the texmf-tree is compiled in 
and so the programs can found the configuration (texmf.cnf) without setting 
any environment variable.  Make sure, that your new pdftex uses the same 
configuration by setting TEXMF or similar (look at the docs).

If you are not so familar with unix (for debugging purposes) I suggest you 
install a up to date version of tetex (the beta from 18.12.2000 works ok 
here and has a new enough pdftex
  pdfTeX, Version 3.14159-14f-released-20000525 (Web2C 7.3.1)

So if you have no solution, examine your system and describe what you have 
done to install (te)tex, pdftex and context.

Yours
Uwe

-- 
mailto:koloska@rcs.urz.tu-dresden.de
http://rcswww.urz.tu-dresden.de/~koloska/
--                                    --
right now the web page is in german only
but this will change as time goes by ;-)

Re: metapost not invoked? mpgraph.1 not created

[Patrick Gundlach]

Hi,

"nothing" happened: is just runs, no error messages are visible. 
(no output file like mpgraph.1 being generated)

gundlach@sarti:~/TeX/context/mprun$ texexec --mptex mpgraph.mp

 TeXExec 2.5 - ConTeXt / PRAGMA ADE 1997-2001

       metapost to tex : mpgraph
            executable : pdfetex
                format : cont-en
             inputfile : tmpgraph
                output : dvips
             interface : en
               options : once
          current mode : all

This is pdfeTeX, Version 3.14159-13d-2.1 (Web2C 7.3.1)
(tmpgraph.tex[/usr/share/texmf/pdftex/config/pdftex.cfg]

ConTeXt  ver: 2001.1.9  fmt: 2001.2.19  int: english  mes: english

language       : language en is active

[....]
Output written on tmpgraph.dvi (1 page, 1128 bytes).
Transcript written on tmpgraph.log.

              run time : 0 seconds
       dvi to metapost : mpgraph
              metapost : mpgraph
                format : metafun
This is MetaPost, Version 0.641 (Web2C 7.3.1)
(mpgraph.mp )
Transcript written on mpgraph.log.

------------------------
mpgraph.log:

This is MetaPost, Version 0.641 (Web2C 7.3.1) (mem=metafun 2001.2.19)  19 FEB 20
01 13:00
**&metafun mpgraph
(mpgraph.mp )
-------------------------
part of tmpgraph.log

This is pdfeTeX, Version 3.14159-13d-2.1 (Web2C 7.3.1) (format=cont-en 2001.2.19
)  19 FEB 2001 16:14
**&cont-en tmpgraph.tex
(tmpgraph.tex[/usr/share/texmf/pdftex/config/pdftex.cfg]

[...]

\openout1 = `mpgraph.mp'.

 (/home/gundlach/TeX/texmf/tex/tufte.tex) [1.1]
systems        : end file tmpgraph at line 4
 ) 

[...]

Output written on tmpgraph.dvi (1 page, 1128 bytes).

--------------------------
the following files changed timestamp/size after the texexec 
--mptex mpgraph.mp run (no files were added to the directory).

changed timestamp:
mpgraph.log, mpgraph.mpx, tmpgraph.tui

changed filesize:
tmpgraph.log, tmpgraph.tmp

---------------------------
the tmpgraph.tmp:
\unprotect
\setupoutput[dvips]
\setupsystem[\c!gebied={/home/gundlach/}]
\protect

----------------------
I have removed the unlink statements in the texexec file. The 
dvi file looks fine, except the output is visible only half. The 
top half is above the upper limit of the 
page. The 
tmpgraph.tex looks like this:

\global \loadfontdefinitionfileoncetrue%
\shipout\hbox{\smash{\hbox{\hbox{% line 84 mpgraph.mp
\framed [width=fit,align=middle]{\input tufte \relax }}\vrule width1sp}}}
\end{document}
\end
------------

the mprun.log file tells me:

mprun.log:systems        : beware: some patches loaded from cont-new.tex!

and

mprun.log:systems        : please update your pdfTeX binaries

anything to worry about for this problem?

this binary is not on my system:
set  DviSpecialScript   to  dvispec

Viele Gruesse,

  Patrick Gundlach

Re: metapost not invoked? mpgraph.1 not created

[Hans Hagen]

At 04:47 PM 2/19/01 +0100, Patrick Gundlach wrote:

>mprun.log:systems        : please update your pdfTeX binaries
>
>anything to worry about for this problem?

eh, maybe, how old is your pdftex? On our site there are pointers to newer
binaries (thanh has a linux one on his site). 

>this binary is not on my system:
>set  DviSpecialScript   to  dvispec

That's ok, since it's unfinished. 

I assume that you run texexec with --pdf ? What happens of you comment the
\...true lines? 

Do others recognize this problem? 

Hans
-------------------------------------------------------------------------
                                  Hans Hagen | PRAGMA ADE | pragma@wxs.nl
                      Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
 tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------

metapost not invoked? mpgraph.1 not created

[Patrick Gundlach]

Hi,

> >\endinput
> 
> no reason for an \endinput here

I had some more stuff after that...

[...]
> >\useMPgraphic{Text}
> >\stoptext
> 
> Goes ok her. Are you using web2c or miktex? 

web2c. tetex 1.0.6 with context installed separately.

this is one output from texexec:

       metapost to tex : mpgraph
            executable : pdfetex
                format : cont-en
             inputfile : tmpgraph
                output : dvips
             interface : en
               options : once batch nomp
          current mode : all

shortly after this:

              run time : 0 seconds
              metapost : mpgraph
                format : metafun
This is MetaPost, Version 0.641 (Web2C 7.3.1)

figures        : figure mpgraph can not be found
[1.1{/usr/share/texmf/dvips/config/pdftex.map

...

this is the tetex part of texexec.ini:

for  tetex  set  UsedInterfaces    to  en,nl,de,uk 
for  tetex  set  UserInterface     to  en 
for  tetex  set  TeXExecutable     to  pdfetex
for  tetex  set  TeXVirginFlag     to  -ini
for  tetex  set  TeXPassString     to  -progname=context 
for  tetex  set  TeXBatchFlag      to  -int=batchmode 
for  tetex  set  MpToTeXExecutable to  mpto
for  tetex  set  MpVirginFlag      to  -ini 
for  tetex  set  MpPassString      to  -progname=mpost
for  tetex  set  MpBatchFlag       to  -int=batchmode 

as far as I remember I have only changed the MpToTeXExecutable 
line and the UserInterface. I have also tried mptotex in the 
MpToTeXExecutable line. (And compiled a mptotex I've found on 
the internet. But this seemed to be even 
worse...)

which perl scripts do I have to use?
I have texexec, mptopdf and texutil in the search path.

after running texexec --pdf mprun.tex  I have the follwoing files
in my directory

mpgraph.log  mpgraph.mpx  mprun.pdf  mprun.tmp  mprun.tuo     tmpgraph.tmp
mpgraph.mp   mprun.log    mprun.tex  mprun.tui  tmpgraph.log  tmpgraph.tui

Viele Gruesse,

  Patrick Gundlach

Re: metapost not invoked? mpgraph.1 not created

[Hans Hagen]

I'll mail you the latest texexec, 

can you run 

  texexec --mptex mpgraph.mp 

and see/report what happens? 

Hans
-------------------------------------------------------------------------
                                  Hans Hagen | PRAGMA ADE | pragma@wxs.nl
                      Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
 tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------

Re: metapost not invoked? mpgraph.1 not created

[Uwe Koloska]

Hello,

just wrote your initial posting again after making my 0,2 Euro suggestions 
for inspecting your system.  Here is another one:

You wrote on Montag, 19. Februar 2001 13:57:
>this is the tetex part of texexec.ini:
>
>for  tetex  set  UsedInterfaces    to  en,nl,de,uk
>for  tetex  set  UserInterface     to  en
>for  tetex  set  TeXExecutable     to  pdfetex
>for  tetex  set  TeXVirginFlag     to  -ini
>for  tetex  set  TeXPassString     to  -progname=context
>for  tetex  set  TeXBatchFlag      to  -int=batchmode
>for  tetex  set  MpToTeXExecutable to  mpto

try to insert the absolute path to mpto here (for me it is 
/opt/teTeX/bin/mpto)  If this works, there is a problem with your path.

And please: try to "answer" the messages because not writing a reply 
disturbs the threading ...
And if answer don't sends to the list, there probably is "answer all" or 
something similar. (don't forget to delete all "To:" entries that don't 
point to the list!)
With kmail2 I can right click on the "To:" field of the displayed message 
and choose "answer to" and get a new message with a reference to the 
displayed one and the text cited the normal way.

Uwe

-- 
mailto:koloska@rcs.urz.tu-dresden.de
http://rcswww.urz.tu-dresden.de/~koloska/
--                                    --
right now the web page is in german only
but this will change as time goes by ;-)

metapost not invoked? mpgraph.1 not created

[Patrick Gundlach]

Hi,

I use the following line to generate my pdf file:
texexec --pdf  rahmen.tex 

I have enabled write18 feature. I have put the following lines in 
rahmen.tex:

\setupcolors[state=start]
\runMPgraphicstrue
\runMPTEXgraphicstrue
\useMETAFUNformattrue
\startuseMPgraphic{Kreis}
  fill fullcircle scaled 200pt withcolor .625yellow ;
\stopuseMPgraphic
\useMPgraphic{Kreis}
\stoptext
\endinput

everything is just fine. After the texexec run, the pdf file includes the 
circle.

But when I have the follwoing tex file:

\setupcolors[state=start]
\runMPgraphicstrue
\runMPTEXgraphicstrue
\useMETAFUNformattrue

\starttext
\startuseMPgraphic{Text}
 draw btex \framed[width=fit,align=middle]{\input tufte \relax} etex ; 
  for i within currentpicture : 
   draw boundingbox i withpen pencircle scaled .2pt withcolor .625yellow ; 
  endfor ;
\stopuseMPgraphic
\useMPgraphic{Text}
\stoptext
\endinput

the mpgraph.1 will not be generated. (And an error appears in the pdf file 
"file mpgraph.1 unknown"). The mpgraph.log tells me:

This is MetaPost, Version 0.641 (Web2C 7.3.1) (mem=metafun 2001.2.17)  17 
FEB 2001 12:42
**&metafun mpgraph
(mpgraph.mp )

when I run 'mpost' and answer '&metafun mpgraph' when mpost outputs the two 
**, the log looks like this:

This is MetaPost, Version 0.641 (Web2C 7.3.1) (mem=metafun 2001.2.17)  17 

FEB 2001 12:42
**&metafun mpgraph
(mpgraph.mp
{randomseed:=2151}
write0 = `mpgraph.mpo'.

 [1] )
1 output file written: mpgraph.1

then the next texexec run (exactly as above) looks correct to me.

Why does texexec (?) not create the mpgraph.1 ?

btw: TeXExec 2.5, pdfeTeX, Version 3.14159-13d-2.1, ConTeXt  ver: 2001.1.9 
-- 
Viele Grüße, 

    Patrick Gundlach

Re: metapost not invoked? mpgraph.1 not created

[Hans Hagen]

At 04:24 PM 2/17/01 +0100, Patrick Gundlach wrote:
>Hi,
>
>
>I use the following line to generate my pdf file:
>texexec --pdf  rahmen.tex 
>
>\useMPgraphic{Kreis}
>\stoptext
>\endinput

no reason for an \endinput here

>everything is just fine. After the texexec run, the pdf file includes the 
>circle.
>
>But when I have the follwoing tex file:
>
>
>\setupcolors[state=start]
>\runMPgraphicstrue
>\runMPTEXgraphicstrue
>\useMETAFUNformattrue
>
>\starttext
>\startuseMPgraphic{Text}
> draw btex \framed[width=fit,align=middle]{\input tufte \relax} etex ; 
>  for i within currentpicture : 
>   draw boundingbox i withpen pencircle scaled .2pt withcolor .625yellow ; 
>  endfor ;
>\stopuseMPgraphic
>\useMPgraphic{Text}
>\stoptext

Goes ok her. Are you using web2c or miktex? 

Hans
-------------------------------------------------------------------------
                                  Hans Hagen | PRAGMA ADE | pragma@wxs.nl
                      Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
 tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------