The FreeType Project - Bugs: bug #55499, Wrong outline orientation of...
You are not allowed to post comments on this tracker with your current authentication level.
bug #55499: Wrong outline orientation of mirrored sub-glyphs
Submitter: | None | ||
Submitted: | Thu 17 Jan 2019 03:47:01 PM UTC | ||
Severity: | 3 - Normal | Item Group: | Incorrect behaviour |
Status: | Confirmed | Privacy: | Public |
Assigned to: | wl | Open/Closed: | Open |
Planned Release: | None |
Mon 28 Sep 2020 07:07:04 AM UTC, comment #3: |
Anonymous |
Sat 19 Jan 2019 04:44:25 PM UTC, comment #2: FT_OUTLINE_REVERSE_FILL should be set for each contour. Then the rasterizer could be updated to work even fro the composite glyphs with opposite orientations. This is not possible without breaking compatibility. Therefore the composite glyphs (and only them) should indeed reverse the point order accordingly.
|
Alexei Podtelezhnikov <podtelez> |
Fri 18 Jan 2019 12:14:32 PM UTC, comment #1: Thanks for the report.
|
Werner LEMBERG <wl> |
Thu 17 Jan 2019 03:47:01 PM UTC, original submission:
Glyph outlines do not always match the FT_OUTLINE_REVERSE_FILL flag.
|
Anonymous |
Depends on the following items: None found
Items that depend on this one: None found
There are 0 votes so far. Votes easily highlight which items people would like to see resolved in priority, independently of the priority of the item set by tracker managers.
Follow 4 latest changes.
Date | Changed by | Updated Field | Previous Value | => | Replaced by |
---|---|---|---|---|---|
2020-09-28 | None | Attached File | - | Added OutlineOrientationTest.ttf, #49875 | |
Attached File | - | Added Right@outline@orientation@of@mirrored@sub-glyphs.PNG, #49876 | |||
2019-01-18 | wl | Status | None | Confirmed | |
Assigned to | None | wl |
Powered by Savane 3.13-caa5.
Corresponding source code
It is verified that reversed references do NOT reverse their point order to compensate. This is verified with OutlineOrientationTest. (16ppem)
The outline does not reverse point order. Right outline orientation of mirrored sub-glyphs.PNG
TD renderer uses FreeType outlines but uses its own rasterizer. The second glyph consists of both the upright reference and a reversed reference. It can be seen that both FreeType and Microsoft outlines have the intersection cancel out, which would not happen if the point order were to be reversed. The FreeType renderer has Half pixel glitching ( https://savannah.nongnu.org/bugs/?58373 ), while TD renderer renders it properly, but the outline itself is correct in all cases, unless the scaled references were to be hinted, then FreeType would have incorrect hinting of scaled subglyphs ( https://savannah.nongnu.org/bugs/?55001 ).
(file #49875,