Description of problem: After selecting Turkish language in GUI install, the text is weird. Like if the wrong font were used. Some chars are printed as a square with 4 digits in it. Version-Release number of selected component (if applicable): Cambridge alpha2 How reproducible: Steps to Reproduce: 1. start GUI install 2. select Turkish language 3. Actual results: Some bad characters Expected results: No bad characters Additional info:
This was fine in Shrike I believe... Owen, do you know of any relevant changes?
What urw-fonts, fontconfig, and FreeType versions?
A2 had fontconfig-2.1-9.i386.rpm fontconfig-devel-2.1-9.i386.rpm freetype-2.1.3-9.i386.rpm freetype-demos-2.1.3-9.i386.rpm freetype-devel-2.1.3-9.i386.rpm freetype-utils-2.1.3-9.i386.rpm urw-fonts-2.0-29.noarch.rpm
Hmm, those are the same versions as in Red Hat 9. Are you sure you didn't remove some fonts from the install image that were there in 9?
Same fonts are present. And still happening with current trees.
Can you provide a string that doesn't display correctly? (perferably as an attachment to avoid bugzilla encoding mangling)
I case that was addressed at me : no. I just started the GUI install in Turkish language and the problems were right there, obvious. Probably on the first page, if not then somewhere in the first 5 pages ( I don't recall anymore ) . I did not perform a complete Turkish install. I don't know turkish.
Well, it was addressed at whoever wanted to dig up the Anaconda .po files, and correlate them with the particular strings that were misdisplaying. Mostly at Jeremy.
*** Bug 106235 has been marked as a duplicate of this bug. ***
Oh, ick. Worked around in rhpl for now. Basic cause is that python's gettext looks at mofiles to find project-id-version. It string.lower()'s the string from the mofile itself and then compares against the lowercase version. With Turkish, this becomes "project-Id-version" (that is, lower("I") = "i") and so no metadata is gathered and then rhpl was falling back to assuming iso8859-1. Sent misa the python patch to look at and changed rhpl.translate to assume UTF-8 instead as that's a safer plan with our current locale strategy.
Mass-closing lots of old bugs which are in MODIFIED (and thus presumed to be fixed). If any of these are still a problem, please reopen or file a new bug against the release which they're occurring in so they can be properly tracked.