Am 24.08.20 um 18:02 schrieb Jean-Marc Lasgouttes: >> After a longer discussion here I found, that the cause of this problem >> was the interaction of LyX with QT5-Qt-5.9.4. Then I compiled LyX with >> QT4 and this workaround was fine for me. >> Now with LyX 2.3.5.2 there is the same problem with QT5-5.12.7 and again > ... > Unfortunately, it is not my failing memory that prevented from fixing > the bug. Since I cannot reproduce it, it is difficult for me to do > something about it. > > Let's start by a shot in the dark: does it make a difference to set the > environment variable QT_HARFBUZZ to "old" ? > > Even if it does, I suspect it will not be enough for a long-term fix. > > Did I understand correctly that you can compile LyX? Does it mean that > you can try a patch too? A possibility is that Qt does not like (ass > Pavel noted) to have a font with a character at position 1 (which is > where \int lives in esint10 font). We would have to duplicate the > character somewhere else in the font. > ...
Hi Jean-Marc, thank you for your kind reply. This morning I tried to compile LyX 2.3.5.2 with your suggested environment-variable export QT_HARFBUZZ=old , and yes, it works, all integral-signs are Ok. > Did I understand correctly that you can compile LyX? Does it mean that > you can try a patch too? Well, I can compile LyX. Concerning an patch I can try - depends if this patch will automatically change the sources, because I don't want to make changes in the sources by hand :-) So far, so good. And many thanks for all your wonderful LyX-work! bernhard -- Homepage: https://www.mb-schiekel.de/ GnuPG: http://p80.pool.sks-keyservers.net/
signature.asc
Description: OpenPGP digital signature
-- lyx-devel mailing list lyx-devel@lists.lyx.org http://lists.lyx.org/mailman/listinfo/lyx-devel
