From Fedora Project Wiki

(moved review stuff to subpage)
No edit summary
Line 7: Line 7:


Taking tentative first steps at getting involved in fonts on Fedora. At the moment, that means resurrecting some of my old ones and learning the new tools.
Taking tentative first steps at getting involved in fonts on Fedora. At the moment, that means resurrecting some of my old ones and learning the new tools.
== FontForge bugs ==
{|
! bug
! FF ver
! description
! actions
|-
| [[rhbug:536920|536920]]
| 20090923-1.fc12
| fontlint crash on neurpoli.ttf from poker3d-data
| reproduced, patched upstream and bz
|-
| [[rhbug:572427|572427]]
| 20090622-2
| cvt window expansion
| build 20100501 works. could be patched for 20090923
|-
| [[rhbug:600108|600108]]
| 20100501-1.fc13
| serafettin ftbfs
| expand stroke issue. reduced, filed upstream. wrong in cvs head (last checked 2010-07-15)
|-
| [[rhbug:604821|604821]]
| 20090923-3.fc13
| crashing after saving to FAT disk and changing name list
| backtrace is corrupt and can't reproduce. resolve:cantfix?
|-
| [[rhbug:608477|608477]]
| 20090923-3.fc13
| Rendering after OKing Grid Fit Parameters dialog crashes in FT_ConicTo()
| Can't reproduce, despite trying perverse settings in dialog. Not a freetype-freeworld issue, as that wasn't installed (so not BCI). Puzzled.
|-
| [[rhbug:609103|609103]]
| 20090923-3.fc13
| _SplineCharLayerFindBounds()
| no info
|-
| [[rhbug:610026|610026]]
| 20090923-3.fc13
| FVMetricsCenter()
| had a go, couldn't reproduce but encountered another bug in same area. filed upstream. TODO: Produce patch for 20090923 and attach
|-
| [[rhbug:610168|610168]]
| 20090923-3.fc13
| edit cvt
| scratch build (20100501) has fixed this
|}
20100501 build from upstream is in Rawhide and a scratch build is available for F13. Currently, the new work on expand stroke in 20100501 causes serafettin-cartoon-fonts to FTBFS, or that would be a good candidate for upgrading F13.
[[rhbug:608477|608477]] is puzzling because the backtrace is clear. Digging in fontforge-devel archives suggests that the Freetype rendering used here has previously (2007-08-05) caused problems because an internal API is being used. Caused problems when going to FT 2.3.5. Now we're on FT 2.3.11, has this bitten again? Seems this was to do with debugging TrueType hinting, so may not apply.


== Bits and bobs ==
== Bits and bobs ==

Revision as of 11:08, 16 July 2010

Paul ‘Flo’ Williams

Taking tentative first steps at getting involved in fonts on Fedora. At the moment, that means resurrecting some of my old ones and learning the new tools.

FontForge bugs

bug FF ver description actions
536920 20090923-1.fc12 fontlint crash on neurpoli.ttf from poker3d-data reproduced, patched upstream and bz
572427 20090622-2 cvt window expansion build 20100501 works. could be patched for 20090923
600108 20100501-1.fc13 serafettin ftbfs expand stroke issue. reduced, filed upstream. wrong in cvs head (last checked 2010-07-15)
604821 20090923-3.fc13 crashing after saving to FAT disk and changing name list backtrace is corrupt and can't reproduce. resolve:cantfix?
608477 20090923-3.fc13 Rendering after OKing Grid Fit Parameters dialog crashes in FT_ConicTo() Can't reproduce, despite trying perverse settings in dialog. Not a freetype-freeworld issue, as that wasn't installed (so not BCI). Puzzled.
609103 20090923-3.fc13 _SplineCharLayerFindBounds() no info
610026 20090923-3.fc13 FVMetricsCenter() had a go, couldn't reproduce but encountered another bug in same area. filed upstream. TODO: Produce patch for 20090923 and attach
610168 20090923-3.fc13 edit cvt scratch build (20100501) has fixed this

20100501 build from upstream is in Rawhide and a scratch build is available for F13. Currently, the new work on expand stroke in 20100501 causes serafettin-cartoon-fonts to FTBFS, or that would be a good candidate for upgrading F13.

608477 is puzzling because the backtrace is clear. Digging in fontforge-devel archives suggests that the Freetype rendering used here has previously (2007-08-05) caused problems because an internal API is being used. Caused problems when going to FT 2.3.5. Now we're on FT 2.3.11, has this bitten again? Seems this was to do with debugging TrueType hinting, so may not apply.

Bits and bobs