test/text-antialias-subpixel.c test/Makefile.am: No longer XFAIL.
test/text-antialias-gray.c test/text-antialias-subpixel.c
Don't turn off metrics hinting, it doesn't make sense.
Update.
Abstract the cap-addition code from the end of the stroke operation into a new _cairo_stroker_add_caps function.
Call the new _cairo_stroker_add_caps at the beginning of every move_to so that we get caps on every subpath and not just the last one.
New function to handle both calling FT_Done_Face on unscaled->face and decrementing font_map->num_open_faces.
Call new _font_map_release_face_lock_held as approporiate.
Assert that (font_map->num_open_faces == 0) when we're done, to help guarantee the bug is fixed.
Don't call FT_Done_Face anymore, instead assert that (unscaled->face == NULL) by the time this function is called.
Prefer TRUE/FALSE as values for cairo_bool_t have_scale.
cairo_image_surface_create should clear image: NOTABUG. The user can clear if desired, and cairo clearing unconditionally would be performance overhead).
cairo_clip is really slow: FIXED/COVERED. Since the time this entry was added we've got rectangular, pixel-alignend clipping optimized. And there's a bug in bugzilla (https://bugs.freedesktop.org/show_bug.cgi?id=4191) for optimizing IN.
self-intersecting path is wrong: COVERED by test/self-intersecting.c.
Cairo is crashing Xnest: NEEDINFO. Is this still happening? Xfake has been working quite well for a very long time now.
font-size=0 in SVG does bad things: FIXED.
Add target "zips" to create zip file based distribution for Win32.
Produce a cairo.def file for distribution by preprocessing the relevant header files and grepping for declarations of functions that start with cairo_. Works for now, will have to modify later if necessary. Use the .def file when linking on Win32.
Check for FT_GlyphSlot_Embolden()
Call FT_GlyphSlot_Embolden for fonts where FC_EMBOLDEN is set but that aren't bold. Add another private flag for this.
Use _cairo_fixed_to_double and _cairo_double_to_fixed when converting between double and 16.16 fixed point. These functions round nicely, and solve some inconsistencies in rendering between my PPC and x86 machines. Thanks to Owen Taylor for suggesting the fix.
The above change does affect linear gradients slightly, but it seems worth it. Update the reference images to follow.
Use pointer comparison for coincident points to ensure a well-defined ordering, and avoid setting discard on both points. This fixes problems with my Mac's implementation of qsort. Final patch by Bertram Felgenhauer.
Update reference images after the change.
Update from gtk-doc CVS, bringing in some minor output changes. (Eliminate the *** from the messages that don't indicate errors.)
Use += for EXTRA_DIST, since gtk-doc.make already defines it.
