Gamma bugs

[Idris Samawi Hamid <ishamid@attbi.com>]

Dear ConTeXtnicians,

In Gamma (ConTeXt compiled with Omega: texexec --make --alone --tex=omega en) it appears that the Omega's otp execution interferes with a number of ConTeXt's 
macros which are not sufficiently insulated. Consider the two files con-test3.tex and type-xxx.tex (cmss is the base font):

%=========con-test3=================
\usetypescriptfile[type-xxx]
\usetypescript[arabtest]
\setupbodyfont[omarab,10pt]

\def\latinunicode{\ocp\BasicLatinTexUni=lat2uni
      \ocp\BasicLatinUniToFont=uni2lat
      \ocplist\BasicLatinOCP=
      \addbeforeocplist 1 \BasicLatinTexUni
      \addbeforeocplist 1 \BasicLatinUniToFont
      \nullocplist
      \pushocplist\BasicLatinOCP}

\def\texencoding{\clearocplists}

 %\latinunicode

\showframe[text]

\starttext%

\showbodyfont%

\showbodyfontenvironment[omarab]

This is a test {\bfc This is a test} {\tfc This is a test} {\tfx This is a
test}

\switchtobodyfont[12pt] This is a test {\tfc This is a test}

\switchtobodyfont[big] This is a test {\tfc This is a test}
\switchtobodyfont[small] This is a test {\tfc This is a test}

\switchtobodyfont[8pt] This is a test {\tfc This is a test}

\switchtobodyfont[10pt] This is a test {\tfc This is a test}

\switchtobodyfont[12pt] This is a test {\tfc This is a test}

\switchtobodyfont[global] This is a test {\tfc This is a test}

\stoptext
%=========type-xxx==================
\starttypescript [naskh] [omega] [name]

   \definefontsynonym [OmegaArabic]      [cmss10]   % [omarab]
   \definefontsynonym [OmegaArabic-Bold] [cmssbx10] % [omarabb]

   \definefontsynonym [Naskh]      [OmegaArabic]
   \definefontsynonym [NaskhBold]  [OmegaArabic-Bold]

\stoptypescript

\starttypescript [naskh] [omega] [size]

   \definebodyfont
     [20.7pt,17.3pt,14.4pt,12pt,11pt,10pt,9pt,8pt,7pt,6pt]
     [rm]
     [tf=Naskh sa 1,
      bf=NaskhBold sa 1]

\stoptypescript

\starttypescript [arabtest]

   \definebodyfontenvironment
     [omarab]
     [default]
     [interlinespace=2.8ex]

  \definetypeface [omarab] [rm] [naskh] [omega] [omega] %[rscale=1.6]

\stoptypescript
===========================
If you leave \latinunicode uncommented then con-test3 will compile like an ordinary ConTeXt job. Uncomment it and funny things happen:

1. Whenever the text size is whatever size \setupbodyfont defaults to, \tfa, \tfb, \tfc, and \tfd will do nothing at all. So in con-test3,

\switchtobodyfont[10pt] This is a test {\tfc This is a test}

will reproduce <This is a test> twice at 10pt. Interestingly, \tfx and \tfxx work fine. If you change the setup to \setupbodyfont[omarab,12pt] then the line 

\switchtobodyfont[12pt] This is a test {\tfc This is a test}

will reproduce <This is a test> twice at 12pt.

This is really weird. I am at a loss at where to start. Where are \tf<variable> defined so that I can look at the \def's and see if I can modify them somehow (though 
ConTeXt's macros are really scary)?

Hans mentioned that omega crashes on his system but I hope he or someone else here can test this and help me find an answer.

2. I mentioned the once before but I was rather unclear then and hope to be more clear and also more general this time:

\showlayout, \showbodyfont, and \showbodyfontenvironment[omarab] do funny things when the otp's are loaded. \showlayout actually causes omega to crash. 
\showgrid does not print the margin numbers (probably because they are isolated from the otp list). I think the solution is to 

a. make sure the macros are isolated from the otp's being executed (although the grid line numbers in \showgrid are an exception here);

b. give an option to these commands for which font to print the chart out in general. For example, \showbodyfontenvironment[omarab][cmr] would mean to print the 
fontenvironment table in cmr. \showbodyfont[omarab][cmr][\myocplist bt] would show sample characters transcribed by `b' and `t' from omarab but print the table and 
verbatim commands in cmr. \myocplist would have to grouped so it affects nothing but the sample characters.

I am rather afraid to mess around with the ConTeXt macros myself. They look rather frightening. Please don't make me do it:-)..

Your help is greatly appreciated!

Best wishes
Idris

==========================
Dr. Idris Samawi Hamid
Depatment of Philosophy
Colorado State University
Fort Collins, CO 80526