I am actually intending to use LaTeX's graphicx package
under plain tex.
And, I am using 'Cyrillic' version of plain tex, called 'cyrtex';
It loads babel etc.
Graphicx package needs some work-around to be loaded. Here is the
sample:
-------------------
\UNPROTECT % this is from plainenc.def
\catcode`\@11\relax
\let\@hash@\relax
\input miniltx
\input graphicx.sty
\catcode`\@12\relax
\PROTECT
\includegraphics{mypicture}
-------------------
Generally, I compiled it by 4 ways:
pdftex -output-format pdf test
pdftex -output-format dvi test
pdftex -fmt cyrtex -output-format dvi test
pdftex -fmt cyrtex -output-format pdf test
3 of them are fine. But the last (cyrtex with pdf output) gives the
error:
-------------------
(/usr/share/texmf-tetex/tex/context/base/supp-pdf.tex
(/usr/share/texmf-tetex/tex/context/base/supp-mis.tex
loading : Context Support Macros / Miscellaneous (2004.10.26)
)
loading : Context Support Macros / PDF (2004.03.26)
! Missing number, treated as zero.
<to be read again>
\let
\gMPa #1->\csname \@@MP 0\number #1
\endcsname
l.2183 \let
\!MPgMPs\gMPs
-------------------
The reason is that the catcode of '!' is 12 instead of 11. It is very
strange.
Here is the minimal example of that behaviour that I can provide:
-------------------
\UNPROTECT % this is from plainenc.def
\catcode`\@11\relax
\let\@hash@\relax
\input miniltx
\input supp-mis
\unprotect % this is from supp-mis.tex
\the\catcode`\! % will get 11
\newcount\Mycount
\the\catcode`\! % will get 11 in plain pdftex, 12 in cyrillic pdftex
\bye
-------------------
So, *only when* using pdftex with cyrtex format, I noticed that
the catcode of '!' changes *right after* \newcount.
Does anybody have any idea on why this happens? I am perplexed.
regards,
Dmitry
because cyrillic pdftex redefines \write:
pdftex \&cyrpdftex \\relax
...
*\show\write
> \write=macro:
->\protect \write .
and \newcount uses \wlog which is defined in terms of \write; so
supp-miss's \unprotect is cancelled by a call to its *own* \protect
command.
so you've a confusion between cyrplain's concept of \protect-ion and
supp-mis's
i would imagine this creates something of an impasse, but it's
obviously easier to hassle vladimir and hans to get these things
playing together than to use something so complicated as latex.
(after all, latex has all this mechanisms in it. to make it possible
for things to work together... and supp-mis is known to work with
latex. even with latex set up to use cyrillic input and to create t2
encoding output, i expect.)
--
Robin Fairbairns, Cambridge
Thank you very much.
I will report this incopatibility beteen cyrtex and supp-mis to
Vladimir Volovich and Hans Hagen.
regards,
Dmitry