puzzling metafun graphics

puzzling metafun graphics

[Thomas A. Schmitz]

This one hast just cost me two hours, and I'm still baffled. The  
story (in shorthand):  was trying to integrate a positional metafun  
graphic into a presentation. I got no output. I tried making a  
minimal example file to reproduce the problem, and sure enough,  
everything went fine. I rebuilt the entire file, by pasting one  
paragraph of the original file at a time, and I got exactly the  
output I wanted. [Drumroll] I deleted the original file and  
everything associated with it [i.e. all those .tui and .tuo etc.],  
saved the new file from my /tmp directory to my working directory,  
and -- I got no output again. diff shows that both files are  
identical. Looking at the logs, there's only one difference I can spot:

< \openout5 = `test-mpgraph.mp'.
---
 > \openout5 = `05.11.08-mpgraph.mp'.
205,206c206,207
<  [MP to PDF] (./test-mpgraph.1) [MP to PDF] (./test-mpgraph.2)
< [MP to PDF] (./test-mpgraph.3)
---
 >  [MP 05.11.08-mpgraph.1] [MP 05.11.08-mpgraph.2]
 > [MP 05.11.08-mpgraph.3]
211c212
< systems         : end file test at line 426
---
 > systems         : end file 05.11.08 at line 426
213c214
< [flush and process test-mpgraph.mp afterwards] )
---
 > [flush and process 05.11.08-mpgraph.mp afterwards] )

The "MP to PDF" part is missing from the file 05.11.08, which might  
explain why I don't see anything in the resulting pdf. But why? Can  
anybody enlighten me why the same file can behave differently? And  
how I can coax ConTeXt into running this MP to PDF thingy? Thanks

Thomas

Re: puzzling metafun graphics

[Taco Hoekwater]

Thomas A. Schmitz wrote:
> working directory,  and -- I got no output again. diff shows that both 
> files are  identical. Looking at the logs, there's only one difference I 
> can spot:

Perhaps it can be that something decides that your file has the
interesting filename extension '11.08-mpgraph.mp' instead of just
'.mp'?

Taco

Re: puzzling metafun graphics

[Hans Hagen]

Thomas A. Schmitz wrote:

> This one hast just cost me two hours, and I'm still baffled. The  
> story (in shorthand):  was trying to integrate a positional metafun  
> graphic into a presentation. I got no output. I tried making a  
> minimal example file to reproduce the problem, and sure enough,  
> everything went fine. I rebuilt the entire file, by pasting one  
> paragraph of the original file at a time, and I got exactly the  
> output I wanted. [Drumroll] I deleted the original file and  
> everything associated with it [i.e. all those .tui and .tuo etc.],  
> saved the new file from my /tmp directory to my working directory,  
> and -- I got no output again. diff shows that both files are  
> identical. Looking at the logs, there's only one difference I can spot:
>
> < \openout5 = `test-mpgraph.mp'.
> ---
> > \openout5 = `05.11.08-mpgraph.mp'.
> 205,206c206,207
> <  [MP to PDF] (./test-mpgraph.1) [MP to PDF] (./test-mpgraph.2)
> < [MP to PDF] (./test-mpgraph.3)
> ---
> >  [MP 05.11.08-mpgraph.1] [MP 05.11.08-mpgraph.2]
> > [MP 05.11.08-mpgraph.3]
> 211c212
> < systems         : end file test at line 426
> ---
> > systems         : end file 05.11.08 at line 426
> 213c214
> < [flush and process test-mpgraph.mp afterwards] )
> ---
> > [flush and process 05.11.08-mpgraph.mp afterwards] )
>
> The "MP to PDF" part is missing from the file 05.11.08, which might  
> explain why I don't see anything in the resulting pdf. But why? Can  
> anybody enlighten me why the same file can behave differently? And  
> how I can coax ConTeXt into running this MP to PDF thingy? Thanks

So the only difference in the source files is the filename? (should not 
make much of a difference because mp conversion is rather stupid)

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: puzzling metafun graphics

[Thomas A. Schmitz]

OK, I'm answering my own post. The plot thickens. It looks like this  
has something to do with file names. It only happens when I have a  
name with a period in it. Thusly, file 05.11.08.tex will not run MP  
to PDF, when I copy the exact same file to 05_11_08.tex, everything  
works as it should.

This may have something to do with the fact (as I just found out)  
that newtexexec doesn't even want to touch these files. I have two  
identical files in a directory, one called 05.10.18.tex, the other  
05_10_18.tex. Here's what happens when I run newtexexec on them:

TeXExec | processing document '05.10.18.tex'
TeXExec | runtime: 0.006379

(and that's it. No output, no action)

TeXExec | processing document '05_10_18.tex'
TeXExec | tex processing method: context
TeXExec | TeX run 1
TeXExec | tex engine: pdfetex
TeXExec | tex format: cont-en
TeXExec | progname: context
This is pdfeTeX, Version 3.141592-1.30.3-2.2 (Web2C 7.5.5)
\write18 enabled.
entering extended mode
(./05_10_18.tex

ConTeXt  ver: 2005.09.08  fmt: 2005.9.9  int: english  mes: english

[snip]

and so on.

Bug or feature? You decide, Hans!

Best

Thomas

On Sep 12, 2005, at 5:45 PM, Thomas A. Schmitz wrote:

> This one hast just cost me two hours, and I'm still baffled. The  
> story (in shorthand):  was trying to integrate a positional metafun  
> graphic into a presentation. I got no output. I tried making a  
> minimal example file to reproduce the problem, and sure enough,  
> everything went fine. I rebuilt the entire file, by pasting one  
> paragraph of the original file at a time, and I got exactly the  
> output I wanted. [Drumroll] I deleted the original file and  
> everything associated with it [i.e. all those .tui and .tuo etc.],  
> saved the new file from my /tmp directory to my working directory,  
> and -- I got no output again. diff shows that both files are  
> identical. Looking at the logs, there's only one difference I can  
> spot:
>
> < \openout5 = `test-mpgraph.mp'.
> ---
> > \openout5 = `05.11.08-mpgraph.mp'.
> 205,206c206,207
> <  [MP to PDF] (./test-mpgraph.1) [MP to PDF] (./test-mpgraph.2)
> < [MP to PDF] (./test-mpgraph.3)
> ---
> >  [MP 05.11.08-mpgraph.1] [MP 05.11.08-mpgraph.2]
> > [MP 05.11.08-mpgraph.3]
> 211c212
> < systems         : end file test at line 426
> ---
> > systems         : end file 05.11.08 at line 426
> 213c214
> < [flush and process test-mpgraph.mp afterwards] )
> ---
> > [flush and process 05.11.08-mpgraph.mp afterwards] )
>
> The "MP to PDF" part is missing from the file 05.11.08, which might  
> explain why I don't see anything in the resulting pdf. But why? Can  
> anybody enlighten me why the same file can behave differently? And  
> how I can coax ConTeXt into running this MP to PDF thingy? Thanks
>
> Thomas
> _______________________________________________
> ntg-context mailing list
> ntg-context@ntg.nl
> http://www.ntg.nl/mailman/listinfo/ntg-context
>

Re: puzzling metafun graphics

[Hans Hagen]

Thomas A. Schmitz wrote:

> OK, I'm answering my own post. The plot thickens. It looks like this  
> has something to do with file names. It only happens when I have a  
> name with a period in it. Thusly, file 05.11.08.tex will not run MP  
> to PDF, when I copy the exact same file to 05_11_08.tex, everything  
> works as it should.
>
> This may have something to do with the fact (as I just found out)  
> that newtexexec doesn't even want to touch these files. I have two  
> identical files in a directory, one called 05.10.18.tex, the other  
> 05_10_18.tex. Here's what happens when I run newtexexec on them:
>
> TeXExec | processing document '05.10.18.tex'
> TeXExec | runtime: 0.006379
>
> (and that's it. No output, no action)
>
> TeXExec | processing document '05_10_18.tex'
> TeXExec | tex processing method: context
> TeXExec | TeX run 1
> TeXExec | tex engine: pdfetex
> TeXExec | tex format: cont-en
> TeXExec | progname: context
> This is pdfeTeX, Version 3.141592-1.30.3-2.2 (Web2C 7.5.5)
> \write18 enabled.
> entering extended mode
> (./05_10_18.tex
>
> ConTeXt  ver: 2005.09.08  fmt: 2005.9.9  int: english  mes: english
>
> [snip]
>
> and so on.
>
> Bug or feature? You decide, Hans!

bug; but does this mean that the old texexec works ok? 

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: puzzling metafun graphics

[Thomas A. Schmitz]

Well, at least the old texexec processes the file. However, the  
problem with metafun graphics I described occurs in the old texexec  
(I'm not at all sure if the two problems are related).

Thomas

On Sep 12, 2005, at 7:38 PM, Hans Hagen wrote:

> bug; but does this mean that the old texexec works ok?
> Hans
>

Re: puzzling metafun graphics

[Hans Hagen]

Thomas A. Schmitz wrote:

> Well, at least the old texexec processes the file. However, the  
> problem with metafun graphics I described occurs in the old texexec  
> (I'm not at all sure if the two problems are related).

can you make me a smal test 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: puzzling metafun graphics

[Adam Lindsay]

Hans Hagen said this at Mon, 12 Sep 2005 19:38:25 +0200:

>Thomas A. Schmitz wrote:
>
>> OK, I'm answering my own post. The plot thickens. It looks like this  
>> has something to do with file names. It only happens when I have a  
>> name with a period in it.
>>
>> Bug or feature? You decide, Hans!
>
>bug; but does this mean that the old texexec works ok? 

Hang on, could this have to do with the following web2c setting?
 allow_multiple_suffixes = f

(as seen in minstall.pdf...)
-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 Adam T. Lindsay, Computing Dept.     atl@comp.lancs.ac.uk
 Lancaster University, InfoLab21        +44(0)1524/510.514
 Lancaster, LA1 4WA, UK             Fax:+44(0)1524/510.492
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Re: puzzling metafun graphics

[Thomas A. Schmitz]

Hmm -- this is getting curioser and curioser... 
I just verified on my linux partition, and I can confirm that 
I don't have the problem there; it's only in OS X. (I'm having 
different problems with my linux install, but that'll be 
another post...). So Adam's educated guess that it must be some
 setting in a config file is more  than probably right. I'll 
keep you posted about this.

On Tue, 2005-09-13 at 10:26 +0100, Adam Lindsay wrote:
> Hans Hagen said this at Mon, 12 Sep 2005 19:38:25 +0200:
> 
> >Thomas A. Schmitz wrote:
> >
> >> OK, I'm answering my own post. The plot thickens. It looks like this  
> >> has something to do with file names. It only happens when I have a  
> >> name with a period in it.
> >>
> >> Bug or feature? You decide, Hans!
> >
> >bug; but does this mean that the old texexec works ok? 
> 
> Hang on, could this have to do with the following web2c setting?
>  allow_multiple_suffixes = f
> 
> (as seen in minstall.pdf...)

Re: puzzling metafun graphics

[Thomas A. Schmitz]

OK, I'm back on OS X and checked:

1. created file 01_01_01.tex with this content:
\starttext

\startuseMPgraphic{circle}

draw fullcircle scaled 10 cm ;

\stopuseMPgraphic

\useMPgraphic{circle}

\stoptext

compiled fine; circle was there.

2. copied file to 01.01.01.tex. Compilation does not give errors, but  
there is no visible output.

3. Treid setting allow_multiple_suffixes = t in texmf.cnf - no  
difference.

Puzzled

Thomas


On Sep 13, 2005, at 11:26 AM, Adam Lindsay wrote:

> Hang on, could this have to do with the following web2c setting?
>  allow_multiple_suffixes = f
>
> (as seen in minstall.pdf...)

Re: puzzling metafun graphics

[Hans Hagen]

Thomas A. Schmitz wrote:

> OK, I'm back on OS X and checked:
>
> 1. created file 01_01_01.tex with this content:
> \starttext
>
> \startuseMPgraphic{circle}
>
> draw fullcircle scaled 10 cm ;
>
> \stopuseMPgraphic
>
> \useMPgraphic{circle}
>
> \stoptext
>
> compiled fine; circle was there.
>
> 2. copied file to 01.01.01.tex. Compilation does not give errors, but  
> there is no visible output.
>
> 3. Treid setting allow_multiple_suffixes = t in texmf.cnf - no  
> difference.
>
> Puzzled
>
> Thomas

it's a tricky thins: 

in order to determine a suffix programs/scripts assume that in 01.01.01 the last 01 is the suffix (there is no rule that tex is the suffix, it's just the built in fallback); 

someprogram 01.01.01
someprogram 01.01.01.tex 

is different. The multiple suffixes always test for a tex file first but that results in two file searches each time (slows down). 

texexec 01.01.01.tex works ok 
texexec 01.01.01 fails 

newtexexec fails in both cases (i will look into that) 

It's kind of tricky to auto append tex since nowadasy we have xml and other variants ..

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