Bug 55935
| Summary: | System font fall back behavior | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Kenichi Ishibashi <bashi> |
| Component: | Text | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED INVALID | ||
| Severity: | Enhancement | CC: | ap, avi, bentfx.com, jshin, mitz, shanestephens, yuzo |
| Priority: | P2 | ||
| Version: | 528+ (Nightly build) | ||
| Hardware: | PC | ||
| OS: | OS X 10.5 | ||
| URL: | http://crbug.com/62435 | ||
Kenichi Ishibashi
See http://crbug.com/62435 in more detail.
In some ports, when a system font fallback is performed to render some Unicode characters such as PILE OF POO (U+1F4A9), WebKit renders the character by using the last resort glyph even if there are fonts which contain appropriate glyph for the character. It might be worth to render the character by using appropriate glyph if it possible. Here is some notes:
- FireFox renders these characters appropriately even if the page doesn't specify font families for the characters.
- As far as I checked, mac/win WebKit ports render the PILE OF POO by using the last resort glyph if system font fallback occurred.
- Chromium linux port renders the character by using appropriate glyph (if it is available).
I'd like WebKit folks' opinion about it.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Avi Drissman
If there exists a valid glyph for a code point, it should be used. It might not fit typographically, but it's better than just having a box.
Skip
Chromium on Linux doesn't properly handle many glyphs.
http://code.google.com/p/chromium/issues/detail?id=90940
Jungshik Shin
I agree with avi on that. Browsers have to hunt down fonts to cover characters if there is a font to cover them instead of prematurely falling back to last resort glyph.