FS#35039 - [fontconfig] non-double matrix element in 10-scale-bitmap-fonts.conf
Attached to Project:
Arch Linux
Opened by oblique (oblique) - Tuesday, 30 April 2013, 16:30 GMT
Last edited by Jan de Groot (JGC) - Tuesday, 08 October 2013, 07:29 GMT
Opened by oblique (oblique) - Tuesday, 30 April 2013, 16:30 GMT
Last edited by Jan de Groot (JGC) - Tuesday, 08 October 2013, 07:29 GMT
|
Details
Description:
I got the following errors: Fontconfig error: "/etc/fonts/conf.d/10-scale-bitmap-fonts.conf", line 70: non-double matrix element Fontconfig error: "/etc/fonts/conf.d/10-scale-bitmap-fonts.conf", line 70: non-double matrix element Fontconfig warning: "/etc/fonts/conf.d/10-scale-bitmap-fonts.conf", line 78: saw unknown, expected number Additional info: * package version(s): 2.10.92-4 In my case I reproduce it with: HandBrakeCLI -i input.mkv -s 1 --subtitle-burned 1 --preset=Normal -o output.mp4 I'm sure there are many ways but I can't think another one for now. |
This task depends upon
Closed by Jan de Groot (JGC)
Tuesday, 08 October 2013, 07:29 GMT
Reason for closing: Not a bug
Additional comments about closing: As upstream says in bugreport: this is not a problem with fontconfig or its configuration, but (non-free, unfixable) software that links static to older versions of fontconfig.
Tuesday, 08 October 2013, 07:29 GMT
Reason for closing: Not a bug
Additional comments about closing: As upstream says in bugreport: this is not a problem with fontconfig or its configuration, but (non-free, unfixable) software that links static to older versions of fontconfig.
I simply don't install it:
sed -e "/10-scale-bitmap-fonts/d" -i conf.d/Makefile*
This bug is similar to
FS#34918When I try to start skype-4.2.0.11-1-x86_64 I get that message and then skype dies after a couple of seconds.
Is there a way to remove these fonts without playing with the source?
The final section in that 10-scale-bitmap-fonts.conf looks invalid.
There's a section that defines "name" and "double" twice (before line 70) and another section that tries to divide by zero (before line 78) by using a "name" tag twice as a symbolic reference and a name. I don't understand how that's supposed to work at all.
I think those well versed in Arch maintenance are much better suited to actually push the request upstream. Personally, I wouldn't know how to make this determination, and where to go for "upstream" reports.
Also, it would be useful if Arch removed the offending file in its own repository until a fix might come from upstream.
https://bugs.freedesktop.org/show_bug.cgi?id=70258