first attempt to run mp

first attempt to run mp

[Pawel Jackowski na Onet]

Hi all!

I've just made a quick test -- running metapost with ConTeXt. The sample
code was just:

    \starttext
     The simplest metapost code sample
    \startMPgraphic
     fill unitsquare scaled 123 withcolor (.2,.3,.4) ;
    \stopMPgraphic
    \stoptext

which looks ok, right? And of course mpost creates nice navy square from
that code. But running

    texexec  --automprun sample.mp

cause a serious system error; ''There is a problem with mpost.exe and bla
bla bla''. Smart and clever' windows XP try to report that problem to
microsoft.
The problem occur just after terminal says

    This is MetaPost,  Version 0.641 (Web2c 7.3.7x)
    (mpgraph.mp

and after that system display an error box.
As manual says I 've made some changes in texmf.cnf:

    openout_any = a
    shell_escape = t
    allow_multiple_suffixes = f

What should I do next to catch 'serious system error' while running metapost
code included in tex code?
Is the problem comonly known?

Thanks, Pawe/l

Re: first attempt to run mp

[Willi Egger]

Hi Pawel

Pawel wrote: At 09:42 07.10.2003, you wrote:
>Hi all!
>
>I've just made a quick test -- running metapost with ConTeXt. The sample
>code was just:
>
>     \starttext
>      The simplest metapost code sample
>     \startMPgraphic
>      fill unitsquare scaled 123 withcolor (.2,.3,.4) ;
>     \stopMPgraphic
>     \stoptext
>
>which looks ok, right? And of course mpost creates nice navy square from
>that code. But running
>
>     texexec  --automprun sample.mp
>
>cause a serious system error; ''There is a problem with mpost.exe and bla
>bla bla''. Smart and clever' windows XP try to report that problem to
>microsoft.
>The problem occur just after terminal says
>
>     This is MetaPost,  Version 0.641 (Web2c 7.3.7x)
>     (mpgraph.mp
>
>and after that system display an error box.
>As manual says I 've made some changes in texmf.cnf:
>
>     openout_any = a
>     shell_escape = t
>     allow_multiple_suffixes = f
>
>What should I do next to catch 'serious system error' while running metapost
>code included in tex code?
>Is the problem comonly known?

I am using Windows XP myself. I never experienced the mentioned problem 
with metapost.
What I miss in your error message is the used format like
    " This is MetaPost, Version 0.641 (Web2c 7.3.7x) (mem=metafun 2003.9.25) "
Did you generate the metafun format and placed it in e.g. the 
TeX\texmf-var\web2c directory, mktexlsr?

Regards Willi 

Re: first attempt to run mp (system error)

[Pawel Jackowski na Onet]

Problem: serious system error while running 'texexec --automprun'


Hi Willi!

> Pawel wrote: At 09:42 07.10.2003, you wrote:
> >Hi all!
> >
> >I've just made a quick test -- running metapost with ConTeXt. The sample
> >code was just:
> >
> >     \starttext
> >      The simplest metapost code sample
> >     \startMPgraphic
> >      fill unitsquare scaled 123 withcolor (.2,.3,.4) ;
> >     \stopMPgraphic
> >     \stoptext
> >
> >which looks ok, right? And of course mpost creates nice navy square from
> >that code. But running
> >
> >     texexec  --automprun sample.mp
> >
> >cause a serious system error; ''There is a problem with mpost.exe and bla
> >bla bla''. Smart and clever' windows XP try to report that problem to
> >microsoft.
> >The problem occur just after terminal says
> >
> >     This is MetaPost,  Version 0.641 (Web2c 7.3.7x)
> >     (mpgraph.mp
> >
> >and after that system display an error box.
> >As manual says I 've made some changes in texmf.cnf:
> >
> >     openout_any = a
> >     shell_escape = t
> >     allow_multiple_suffixes = f
> >
> >What should I do next to catch 'serious system error' while running
metapost
> >code included in tex code?
> >Is the problem comonly known?
>
> I am using Windows XP myself. I never experienced the mentioned problem
> with metapost.
> What I miss in your error message is the used format like
>     " This is MetaPost, Version 0.641 (Web2c 7.3.7x) (mem=metafun
2003.9.25) "
> Did you generate the metafun format and placed it in e.g. the
> TeX\texmf-var\web2c directory, mktexlsr?

I have format properly generated, so that is not the problem. Even if I
there weren't metafun format 'serious system error' shouldn't occur... So I
have still no idea what is wrong. My friend use similar TeX installation
(based on TeXLive7) and has the same problem. Any ideas?


Pawe/l

Re: first attempt to run mp (system error)

[Willi Egger]

Hi Pawel,

>Pawel wrote:
>Problem: serious system error while running 'texexec --automprun'
>
>
>Hi Willi!
>
> > Pawel wrote: At 09:42 07.10.2003, you wrote:
> > >Hi all!
> > >
> > >I've just made a quick test -- running metapost with ConTeXt. The sample
> > >code was just:
> > >
> > >     \starttext
> > >      The simplest metapost code sample
> > >     \startMPgraphic
> > >      fill unitsquare scaled 123 withcolor (.2,.3,.4) ;
> > >     \stopMPgraphic
> > >     \stoptext
> > >
> > >which looks ok, right? And of course mpost creates nice navy square from
> > >that code. But running
> > >
> > >     texexec  --automprun sample.mp
> > >
> > >cause a serious system error; ''There is a problem with mpost.exe and bla
> > >bla bla''. Smart and clever' windows XP try to report that problem to
> > >microsoft.
> > >The problem occur just after terminal says
> > >
> > >     This is MetaPost,  Version 0.641 (Web2c 7.3.7x)
> > >     (mpgraph.mp
> > >
> > >and after that system display an error box.
> > >As manual says I 've made some changes in texmf.cnf:
> > >
> > >     openout_any = a
> > >     shell_escape = t
> > >     allow_multiple_suffixes = f
> > >
> > >What should I do next to catch 'serious system error' while running
>metapost
> > >code included in tex code?
> > >Is the problem comonly known?
> >
> > I am using Windows XP myself. I never experienced the mentioned problem
> > with metapost.
> > What I miss in your error message is the used format like
> >     " This is MetaPost, Version 0.641 (Web2c 7.3.7x) (mem=metafun
>2003.9.25) "
> > Did you generate the metafun format and placed it in e.g. the
> > TeX\texmf-var\web2c directory, mktexlsr?
>
>I have format properly generated, so that is not the problem. Even if I
>there weren't metafun format 'serious system error' shouldn't occur... So I
>have still no idea what is wrong. My friend use similar TeX installation
>(based on TeXLive7) and has the same problem. Any ideas?

I have really no clue, my metapost.exe  and metapost.dll are from the 
distribution TeXlive7

May be this should also be asked to Fabrice Popineau. Unfortunately I lost 
a lot of mail-addresses due to a crash, so I can't hint you on his addres.

Willi

Re: first attempt to run mp

[Patrick Gundlach]

Hi,

I am not a windows user, but what happens if you run mpost.exe on the
mpgraph.mp file? Do you get the same error?

> As manual says I 've made some changes in texmf.cnf:
>
>     openout_any = a
>     shell_escape = t
>     allow_multiple_suffixes = f

Looks good. 

> What should I do next to catch 'serious system error' while running metapost
> code included in tex code?
> Is the problem comonly known?

I don't think so, but I don't really know.


Patrick
-- 
You are your own rainbow!

Re: Re: first attempt to run mp (system error)

[Paweł Jackowski (onet)]

Problem: system error while running texexec --automprun

07-10-2003, 18:02, Patrick Gundlach: 
> Hi,
> 
> I am not a windows user, but what happens if you run mpost.exe on the
> mpgraph.mp file? Do you get the same error?

Yes, indeed! But compiling code

	beginfig(100)
	 fill unitsquare scaled 123 withcolor (.2,.3,.4);
	endfig;
	end

doesn't causes any errors, which means that so mpost program works ok.
The problem is in the mpgraph.mp. I use the same texmf tree for linux
and windows, so I can compare how it works with linux binaries. En error
also occur in linux:

	This is MetaPost, Version 0.641 (Web2C 7.3.7x)
	(mpgraph.mp (/mnt/winXP/TeX/texmf/metapost/context/mp-tool.mp)
	(/mnt/winXP/TeX/texmf/metapost/context/mp-spec.mp)
	(/mnt/winXP/TeX/texmf/metapost/context/mp-grph.mp)
	(/mnt/winXP/TeX/texmf/metapost/context/mp-page.mp)
	(/mnt/winXP/TeX/texmf/metapost/context/mp-core.mp)Naruszenie 	ochrony
pamięci


The last alert means "Memory violating". A remedy for such problems is
to regenerate the format. However the wrong metafun format was generated
with fmtutil and its default settings... Anyway I have regenerated
metafun format with command:

	texexec --make --alone metafun


But unfortunately regenerating format is not the whole story.
After that I tried to compile under windows following code:

	\starttext
		The simplest metapost code
	\startMPgraphic
		fill unitsquare scaled 123 withcolor (.2,.3,.4);
	\stopMPgraphic
	\stoptext

(Is there something wrong in it?)
After running 'texexec --automprun --pdf sample.tex' MetaPost generated
nice mpgraph.1. I also tried 'mpost mpgraph.mp' which doesn't cause
problems any more. The only problem is that PDF output didn't contain
the graphic!

Under linux the problem is even stranger. Running 'texexec --automprun
--pdf sample.tex' produce (as in windows) proper postscript code
mpgraph.1 (properly interpreted and displayed by ghostscript). But
running 'mpost mpgraph.mp' under linux still causes the "Memory
violating" error. PDF output also doesn't contain the graphic.

Unfortunately mpgraph.log is empty (zero length file) and tex log file
doesn't inform about any problems.

I'm afraid there are at least two independent problems: no graphic in
PDF output and "Memory violating" error (regenerating format can't help
here).


:-?


Thanks, Paawe/l

Re: Re: first attempt to run mp (system error)

[Hans Hagen]

At 09:27 08/10/2003, you wrote:

>Yes, indeed! But compiling code
>
>         beginfig(100)
>         fill unitsquare scaled 123 withcolor (.2,.3,.4);
>         endfig;
>         end

\starttext \startMPpage
...
\stopMPpage \stoptext

The last alert means "Memory violating". A remedy for such problems is
>to regenerate the format. However the wrong metafun format was generated
>with fmtutil and its default settings... Anyway I have regenerated
>metafun format with command:

are you on windows or on linux?

>         texexec --make --alone metafun
>
>Under linux the problem is even stranger. Running 'texexec --automprun
>--pdf sample.tex' produce (as in windows) proper postscript code
>mpgraph.1 (properly interpreted and displayed by ghostscript). But
>running 'mpost mpgraph.mp' under linux still causes the "Memory
>violating" error. PDF output also doesn't contain the graphic.
>
>Unfortunately mpgraph.log is empty (zero length file) and tex log file
>doesn't inform about any problems.
>
>I'm afraid there are at least two independent problems: no graphic in
>PDF output and "Memory violating" error (regenerating format can't help
>here).

ok, take one of:

www.pragma-ade.com/context/mswintex.zip
www.pragma-ade.com/context/linuxtex.zip

(some 20Meg)

this is what's needed for running context

in there you'll find

setuptex
setuptex.bat

on linux, make sure that you "chmod -R 755 *" the tree first

the windows file has formats, in it (i'll remove them some day) but for 
linux you need to generate them.

If all goes well, you'll have a system running now.

(this one has tl8 binaries)

Later i'll put the tree's on the site so that people can mirror them

Hans

first attempt to run mp (system error) -- summary?

[Pawel Jackowski na Onet]

Pawe/l wrote about:
1. System error while compiling nested mp code (under Windows) --
regenerating metafun format solves that
2. 'Memory violating' error while compiling nested mp code (under Linux) --
still can't find the answer...
3. Missed mp graphics in output PDF -- solved of course... sorry for that
silly question, I've taken the first example appearing in ConTeXt manual
assuming that it is the basic one.


>From: "Patrick Gundlach" <pg@levana.de>
>Sent: Wednesday, October 08, 2003 6:32 PM
> you don't need --automprun. Everything should work fine w/o.

I tried to omit this option, but without that MetaPost doesn't run at all.

>From: "Hans Hagen" <pragma@wxs.nl>
>Sent: Wednesday, October 08, 2003 9:53 AM
> ok, take one of:
>
> www.pragma-ade.com/context/mswintex.zip
> www.pragma-ade.com/context/linuxtex.zip
>
> (some 20Meg)

 Thanks a lot! downloading started.

> this is what's needed for running context
>
> in there you'll find
>
> setuptex
> setuptex.bat
>
> on linux, make sure that you "chmod -R 755 *" the tree first
>
> the windows file has formats, in it (i'll remove them some day) but for
> linux you need to generate them.

Doesn't hurt. ...But wait -- I aks to be on the safe side: since I remember,
I use common formats (and every other files in texmf tree) for linux and
windows. Is there ANY reason to generate nativ format for each operating
system? Is there any reason not to make common any other texmf files?

> If all goes well, you'll have a system running now.
> (this one has tl8 binaries)

Surely I will! Which version has tl2003 binarises: linux, windows or both?
By the way, does anybody know if tl2003 binaries are frozen already?

Thanks to all of You!
Best regards,  Pawe/l

Re: first attempt to run mp (system error) -- summary?

[Patrick Gundlach]

Hi,

> Pawe/l wrote about:
> 1. System error while compiling nested mp code (under Windows) --
> regenerating metafun format solves that

good to hear.

> 2. 'Memory violating' error while compiling nested mp code (under Linux) --
> still can't find the answer...

that is really strange. If you are brave and experienced, you might
trace the mpost run with strace to see where the error occurs. 

Please post the logfile of the bad run.

> 3. Missed mp graphics in output PDF -- solved of course... sorry for that
> silly question, I've taken the first example appearing in ConTeXt manual
> assuming that it is the basic one.

This question was not silly at all. I am still confused by the
plethora of ways to run mp.

>> you don't need --automprun. Everything should work fine w/o.
>
> I tried to omit this option, but without that MetaPost doesn't run at all.

could you post the output of texexec --verbose and the logfile of a
simple ConTeXt file with mp code in it?

> Doesn't hurt. ...But wait -- I aks to be on the safe side: since I remember,
> I use common formats (and every other files in texmf tree) for linux and
> windows. Is there ANY reason to generate nativ format for each operating
> system? Is there any reason not to make common any other texmf
> files?

I'd like to know, too.

> Surely I will! Which version has tl2003 binarises: linux, windows or both?
> By the way, does anybody know if tl2003 binaries are frozen already?

<url:http://www.tug.org/texlive/>

For TeX Live 2003, the last update of packages and programs was made
on 20 September 2003.  For the latest versions, please consult CTAN .

Patrick
-- 
You are your own rainbow!

Re: Re: first attempt to run mp (system error) -- summary?

[Pawel Jackowski na Onet]

Problem: memory violoating error under Linux while running mp

From: "Patrick Gundlach" <pg@levana.de>
> [...]
> > 2. 'Memory violating' error while compiling nested mp code (under
Linux) --
> > still can't find the answer...
>
> that is really strange. If you are brave and experienced, you might
> trace the mpost run with strace to see where the error occurs.
>
> Please post the logfile of the bad run.

As I mentioned, created 'mpgraph.log' file has zero length...

 > >> you don't need --automprun. Everything should work fine w/o.
> >
> > I tried to omit this option, but without that MetaPost doesn't run at
all.
>
> could you post the output of texexec --verbose and the logfile of a
> simple ConTeXt file with mp code in it?

Ahh, to late! Finally I decided to remove tl7 installation (just before i
recived your answer) and try to make an upgrade. So if the problem will
occur again, I will report it to you.


Thanks, Pawe/l

Re: first attempt to run mp (system error) -- summary?

[Patrick Gundlach]

Hi again,

...

> Doesn't hurt. ...But wait -- I aks to be on the safe side: since I remember,
> I use common formats (and every other files in texmf tree) for linux and
> windows. Is there ANY reason to generate nativ format for each operating
> system? Is there any reason not to make common any other texmf
> files?

If you have the same system on both sides, I think you are safe. It
looks as if web2c systems can share formats if the web2c version
number is identical. 

Patrick
-- 
You are your own rainbow!

Re: Re: first attempt to run mp (system error) -- summary?

[Hans Hagen]

At 23:41 08/10/2003, you wrote:
>Hi again,
>
>...
>
> > Doesn't hurt. ...But wait -- I aks to be on the safe side: since I 
> remember,
> > I use common formats (and every other files in texmf tree) for linux and
> > windows. Is there ANY reason to generate nativ format for each operating
> > system? Is there any reason not to make common any other texmf
> > files?
>
>If you have the same system on both sides, I think you are safe. It
>looks as if web2c systems can share formats if the web2c version
>number is identical.

indeed and this is what makes tex live such a nice system

however, here i organize my binaries like

texmf-mswin
texmf-linux

etc because i often have more modern pdftex's on windows -)

Hans  

Re: first attempt to run mp (system error) -- summary?

[Henning Hraban Ramm]

Am Mittwoch, 08.10.03, um 22:19 Uhr (Europe/Zurich) schrieb Pawel 
Jackowski na Onet:
>> From: "Patrick Gundlach" <pg@levana.de>
>> Sent: Wednesday, October 08, 2003 6:32 PM
>> you don't need --automprun. Everything should work fine w/o.
>
> I tried to omit this option, but without that MetaPost doesn't run at 
> all.

Did you uncomment the MP entries in cont-usr.tex?
(That is, copy cont-usr.ori to cont-usr.tex and change that.)
After that you must re-generate the formats and run texhash (=mktexlsr).

Grüßlis vom Hraban!
-- 
http://www.fiee.net/texnique/

Re: Re: first attempt to run mp (system error)

[Willi Egger]

At 09:27 08.10.2003, Pawel wrote:
>Problem: system error while running texexec --automprun
>
>07-10-2003, 18:02, Patrick Gundlach:
> > Hi,
> >
> > I am not a windows user, but what happens if you run mpost.exe on the
> > mpgraph.mp file? Do you get the same error?
>
>Yes, indeed! But compiling code
>
>         beginfig(100)
>         fill unitsquare scaled 123 withcolor (.2,.3,.4);
>         endfig;
>         end
>
>doesn't causes any errors, which means that so mpost program works ok.
>The problem is in the mpgraph.mp. I use the same texmf tree for linux
>and windows, so I can compare how it works with linux binaries. En error
>also occur in linux:
>
>         This is MetaPost, Version 0.641 (Web2C 7.3.7x)
>         (mpgraph.mp (/mnt/winXP/TeX/texmf/metapost/context/mp-tool.mp)
>         (/mnt/winXP/TeX/texmf/metapost/context/mp-spec.mp)
>         (/mnt/winXP/TeX/texmf/metapost/context/mp-grph.mp)
>         (/mnt/winXP/TeX/texmf/metapost/context/mp-page.mp)
>         (/mnt/winXP/TeX/texmf/metapost/context/mp-core.mp)Naruszenie 
> ochrony
>pamiêci
>
>
>The last alert means "Memory violating". A remedy for such problems is
>to regenerate the format. However the wrong metafun format was generated
>with fmtutil and its default settings... Anyway I have regenerated
>metafun format with command:
>
>         texexec --make --alone metafun
>
>
>But unfortunately regenerating format is not the whole story.
>After that I tried to compile under windows following code:
>
>         \starttext
>                 The simplest metapost code
>         \startMPgraphic
>                 fill unitsquare scaled 123 withcolor (.2,.3,.4);
>         \stopMPgraphic
>         \stoptext
>
>(Is there something wrong in it?)
>After running 'texexec --automprun --pdf sample.tex' MetaPost generated
>nice mpgraph.1. I also tried 'mpost mpgraph.mp' which doesn't cause
>problems any more. The only problem is that PDF output didn't contain
>the graphic!

Shure, the graphic was not called to be placed in the document!
So you might add  \useMPgraphic
This works here:

\setupoutput[pdftex]
\setupcolors[state=start]

\starttext
                 The simplest metapost code
         \startuseMPgraphic{test}
                 fill unitsquare scaled 123 withcolor (.2,.3,.4);
         \stopuseMPgraphic
              \useMPgraphic{test}
\stoptext

Willi

>Under linux the problem is even stranger. Running 'texexec --automprun
>--pdf sample.tex' produce (as in windows) proper postscript code
>mpgraph.1 (properly interpreted and displayed by ghostscript). But
>running 'mpost mpgraph.mp' under linux still causes the "Memory
>violating" error. PDF output also doesn't contain the graphic.
>
>Unfortunately mpgraph.log is empty (zero length file) and tex log file
>doesn't inform about any problems.
>
>I'm afraid there are at least two independent problems: no graphic in
>PDF output and "Memory violating" error (regenerating format can't help
>here).
>
>
>:-?
>
>
>Thanks, Paawe/l
>
>
>_______________________________________________
>ntg-context mailing list
>ntg-context@ntg.nl
>http://www.ntg.nl/mailman/listinfo/ntg-context

Re: first attempt to run mp (system error)

[Patrick Gundlach]

Hi,

> 	\starttext
> 		The simplest metapost code
> 	\startMPgraphic
> 		fill unitsquare scaled 123 withcolor (.2,.3,.4);
> 	\stopMPgraphic
> 	\stoptext

there is also \startMPcode....\stopMPcode

> After running 'texexec --automprun --pdf sample.tex' MetaPost
> generated

you don't need --automprun. Everything should work fine w/o.

Patrick
-- 
You are your own rainbow!