Dear all I appear to have run up against a bug discussed last March. I'm newly running: mtx-context | main context file: /Users/jjc/context11-11/tex/texmf-context/tex/context/base/context.mkiv mtx-context | current version: 2011.11.14 16:17 with a short test file I run: context --environment=arabtestI.tex --result=tmp arab-testII.xml and the log concludes thus: mtx-context | main context file: /Users/jjc/context11-11/tex/texmf-context/tex/context/base/context.mkiv mtx-context | current version: 2011.11.14 16:17 but when I try to do the same thing on my full marked up text: context --environment=svo-tei.tex --result=tmp svoboda_diary.xml the log concludes like this: mtx-context | fatal error: return code: 1 I didn't notice this until I tried to use the --result flag because, in spite of the 'fatal error', a new pdf (that actually reflected changes I had made in the environment file) was in fact created using the name of the xml input: svoboda_diary.pdf. But when I used --result, a new pdf was created for the short file but not for the long one. Wondering why, I inspected the log and only then noticed: "mtx-context | fatal error: return code: 1" In the error case the lua stats in the log shows: mkiv lua stats > result saved in file - tmp.pdf but no tmp.pdf file is produced. Instead, svoboda_diary.pdf is updated; weird huh? In the discussion on the list Steffan said "Taco found the bug (thanks!): There is a fix luatex rev. 4093." So am I seeing a regression error here or, as is perfectly possible since I'm out there beyond the limits of my understanding, am I just doing something wrong or using the wrong tool? baffled as usual, Jon