Am Mon, 20 May 2024 10:40:35 -0400 schrieb Scott Kostyshak <skost...@lyx.org>:
> On Mon, May 20, 2024 at 03:11:41PM GMT, Kornel Benko wrote: > > Am Mon, 20 May 2024 14:27:26 +0200 > > schrieb "Jürgen Spitzmüller" <jspi...@gmail.com>: > > > > > Am Montag, dem 20.05.2024 um 11:40 +0200 schrieb Kornel Benko: > > > > Some russian documents fail > > > > > > They work for me if I set appropriate fonts. What error do you get? > > > > After rerunning the Russian tests, they passed here too. Interestingly the > > needed time > > was way smaller (instead of about 1000s now they needed each 30-60s). > > The problem with those tests is that, from the little that I understand, > LuaTeX does not generate the fonts needed. The reason they pass on your > second try is that other engines do generate the fonts. > > In lyx-tester I build the fonts manually before running the tests. It is > hackish, but you can see it here: > > > https://gitlab.com/scottkosty/lyx-tester/-/blob/master/lyx-tester?ref_type=heads#L561 > Probably a better hack would be to run the other tests first. e.g., we > could add some dependencies like "pdf5 and dvi3 test for Russian > documents must be run after the pdf2 tests", but I don't know if that's > possible with the ctests. I don't see a way to order the tests. Maybe run first run xelatex tests with '-R pdf4_systemF' and then run with added '-E pdf4_systemF' > I asked a question for alternative workarounds: > > > https://tex.stackexchange.com/questions/649259/can-i-ask-luatex-to-run-mktextfm Not much help there :( > By the way, this issue had stumped me for probably about a year before I > figured it out, because it seemed random and not reproducible. > > Scott Kornel
pgp3LnIvm7L24.pgp
Description: Digitale Signatur von OpenPGP
-- lyx-devel mailing list lyx-devel@lists.lyx.org http://lists.lyx.org/mailman/listinfo/lyx-devel
