I noticed recently, that all programs that are using fontconfig print this on startup: > Fontconfig error: failed reading config file > Fontconfig error: Cannot load config file "infinality/conf.d" This problem was reported on the forum with a proposed solution in eselect-infinality: https://forums.gentoo.org/viewtopic-p-8205494.html I'm not sure which fontconfig version introduced this, but think we should fix it.
Created attachment 541816 [details, diff] Fix eselect-infinality to create a symlink to an absolute path Thanks for reporting this. I just stumbled over the forums thread you linked to, when I had tons of errors when running doxygen. The attached patch is a simple fix (one-liner): eselect-infinality must *NOT* create a relative symlink , but link /etc/fonts/infinality/conf.d to an absolute path in /etc/fonts/infinality/styles.conf.avail/
Still waiting for response from fonts team...
(In reply to Maxim Koltsov from comment #2) > Still waiting for response from fonts team... Sorry, the fonts team lacks manpower a bit. I will have a look at a fix for this today.
Also, it might be related to this as well; the error message is the same: https://bugs.gentoo.org/650248 THis is an old bug that has just today become rather important for infinality users, because chromium now depends on =>fontconfig-2.13.0-r4, which completely breaks all ionfinality font settings. So far, I've survived by masking all versions of fontconfig-2.13.0, but now am not sure what to do with this new dependency.
BTW, 2.12.6 is the last version NOT to display that error message, and it's the last version to respect infinality configuration settings.
The bug has been closed via the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=2c53799d22cecefea1b320a6b964bb957351327b commit 2c53799d22cecefea1b320a6b964bb957351327b Author: Guilherme Amadio <amadio@gentoo.org> AuthorDate: 2018-07-31 17:44:16 +0000 Commit: Guilherme Amadio <amadio@gentoo.org> CommitDate: 2018-07-31 17:44:16 +0000 app-eselect/eselect-infinality: version bump to 1.1 Closes: https://bugs.gentoo.org/659772 Package-Manager: Portage-2.3.44, Repoman-2.3.10 app-eselect/eselect-infinality/Manifest | 1 + .../eselect-infinality-1.1.ebuild | 27 ++++++++++++++++++++++ 2 files changed, 28 insertions(+)
I adopted eselect-infinality since the former maintainer Ben de Groot (yngwin) has retired. Please let me know if something is still not working. We can ask for stabilization sooner if fontconfig-2.13 is marked stable.
I don't think this is resolved; the error message still exists on my setup, and the fonts are not correct. I've removed my local masks for fontconfig and pango, and rebuilt them to the latest versions, as well as eselect-infinality. The error still exists on bootup: > Fontconfig error: failed reading config file > Fontconfig error: Cannot load config file "infinality/conf.d" I don't know about other users, but it isn't solved for me, and it does not solve https://bugs.gentoo.org/650248 either. Am I missing something?
(In reply to thunderrd from comment #8) > I don't think this is resolved; the error message still exists on my setup, > and the fonts are not correct. > > I've removed my local masks for fontconfig and pango, and rebuilt them to > the latest versions, as well as eselect-infinality. The error still exists > on bootup: > > > Fontconfig error: failed reading config file > > Fontconfig error: Cannot load config file "infinality/conf.d" > > I don't know about other users, but it isn't solved for me, and it does not > solve https://bugs.gentoo.org/650248 either. > > Am I missing something? Please rerun eselect infinality and set the configuration to the same thing. The links are not reset by just reinstalling eselect-infinality. New users will not get a faulty infinality when installing the newer version.
Thank you; you are absolutely correct. Resetting the links again seems to have solved the problem. Perhaps you can make a comment in https://bugs.gentoo.org/650248 to mention the need for the reset? Then I think you can close that bug as well.
*** Bug 650248 has been marked as a duplicate of this bug. ***