[Please cc Alan for this thread.] Hans Hagen wrote:
features, and a good torture file would help immensely in noticing these things. I'll try to whip something up in addition to Omega's torture.tex (which I could never get to work anyway...)
forget about such tests, the real test for eomega is a 300 page document with average memory settings; the main thing is to make sure that the otp parser does not keep on reading whole files
I have some pretty intense stuff here, but unfortunately the otp's and fonts are proprietary so we can't use them for general testing. But I have some all-vowelized Arabic stuff extending over a hundred pages that generally works fine. Some of it uses EDMAC, which really helps to tax the memory etc. Fully vowelized Arabic is the most taxing stuff; I guess I can port some of my stuff to the default Omega font (omarb) for testing, though the results will look chaotic unless I write new otp's for that font (ugh). On the other hand, since u and gb don't know Arabic :-( that will not matter at all. I'll try to put something together in the next two weeks (grading, teaching Summer session, going to DC for a speech r keeping me tied up 4 now). Best Idris