On Nov 23, 2007 11:53 AM, Steffen Wolfrum wrote:
As this strange error makes me a bit nervous I checked an older file for its behavior: same result!
A file that without problems was typeset in April now ends up with the same kind of freezing:
\usetypescriptfile[type-gyr]
you don't need that one for XeTeX, I think.
\usetypescript[times][ec]
It's still not clear to me how to make these definition engine-independent, but for XeTeX use \usetypescript[times] else you'll get the old ec-encoded fonts instead of OpenType.
\mainlanguage[de] \enableregime[utf]
\setupbodyfont[times,10pt] \setupinterlinespace[line=12pt]
\starttext
Wort und Tat in 1906
\stoptext
When this is set with pdftext all is fine. With xetex (or xtx) the process stop at this line:
(/usr/local/texlive/2007/../texmf-local/tex/context/base/type- gyr.tex))! unknown encoding "utf"; reading as raw bytes
Sorry, my fault again. For you: to solve the problem temporary, remove \enableregime, or use \beginOLDTEX \enableregime[utf] \endOLDTEX or try to fix this line in regi-ini.mkii (and test the behavior): \def\mkenableregime#1{\doifnot{#1}{utf}{\XeTeXinputencoding "#1"\relax}} For Hans: There's a little problem in regi-ini: \def\mkenableregime#1{\XeTeXinputencoding "#1"\relax} \def\enableregime[#1]% {\edef\currentregime{\trueregimename{#1}}% \doifelsenothing\currentregime {\disableregime} {\douseregime\currentregime \mkenableregime\currentregime}} This passes \XeTeXinputencoding "utf" to XeTeX, even is someone says \enableregime[utf-8], but XeTeX doesn't know what "utf" encoding is (it expects at least "utf-8" or nothing). The following fix helps, but is not the final answer: \def\mkenableregime#1{\doifnot{#1}{utf}{\XeTeXinputencoding "#1"\relax}} This has one weak point though: \enableregime[cp1250] \enableregime[utf-8] will leave the current regime at cp1250, so perhaps \def\mkenableregime#1{\doifelse{#1}{utf}% {\XeTeXinputencoding "utf-8"\relax}{\XeTeXinputencoding "#1"\relax}} Another option would be to rename utf to utf-8 internally (or to adapt \trueregimename or regi-syn), but that would probably do more harm than good unless tested well, and it's not worth the effort. ... Hmmm .... when thinking agin ... I know that you asked me which regime name to pass to XeTeX and I said "don't worry, it should be more or less the same", but "8859-1" isn't valid regime name either. I'll take a look and report back. Mojca