| Summary: | Changes to freetype module break optimization heuristics on CJK fonts. | ||||||
|---|---|---|---|---|---|---|---|
| Product: | xorg | Reporter: | Egbert Eich <eich> | ||||
| Component: | Server/General | Assignee: | Roland Mainz <roland.mainz> | ||||
| Status: | RESOLVED FIXED | QA Contact: | |||||
| Severity: | normal | ||||||
| Priority: | high | CC: | ajax, cyamauch, roland.mainz, xorg-team | ||||
| Version: | unspecified | ||||||
| Hardware: | x86 (IA32) | ||||||
| OS: | Linux (All) | ||||||
| URL: | http://freedesktop.org/pipermail/release-wranglers/2004-March/000240.html | ||||||
| Whiteboard: | |||||||
| i915 platform: | i915 features: | ||||||
| Bug Depends on: | 556 | ||||||
| Bug Blocks: | |||||||
| Attachments: |
|
||||||
.... to be removed once a better solution becomes available. Taking bug. In the meantime the tree has updated to Freetype2.1.8 (that was bug 556) which include a FreeType API extension which allows that we can fix this bug while using only "public" APIs. Created attachment 260 [details] [review] Patch for 2004-05-04-trunk The patch is the same as Chisato's original patch at http://x-tt.sourceforge.jp/arch/X11R6.7.0-libfreetype-xtt2-1.2a.patch Patch checked-in... Checking in xc/ChangeLog; /cvs/xorg/xc/ChangeLog,v <-- ChangeLog new revision: 1.14; previous revision: 1.13 done Checking in xc/lib/font/FreeType/ftfuncs.c; /cvs/xorg/xc/lib/font/FreeType/ftfuncs.c,v <-- ftfuncs.c new revision: 1.3; previous revision: 1.2 done Checking in xc/lib/font/FreeType/ftfuncs.h; /cvs/xorg/xc/lib/font/FreeType/ftfuncs.h,v <-- ftfuncs.h new revision: 1.3; previous revision: 1.2 done Mailing the commit message to xorg-commit@pdx.freedesktop.org... .. marking bug as FIXED. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.
The above URL points to an email thread discussing this issue. Keith Packard suggested to pull a single glyph from the font to find out if it gives back a bitmap or outline and base the decision wether to use the 'very lazy' method based on this (under the assumption that if there is a bitmap for one glyph there are bitmaps for all. I would like to propose the following procedure: - Implement the proposal from above and test it. If the results are bad: - Turn to the freetype people to add a way to obtain the required information thru the standard API. - For the short term reenable the old code for those who compile the .a module using the builtin lib. Clearly mark this as a hack that will have