165878 - Only the first one of the suggested fonts is passed to ATSUI

Status: RESOLVED WORKSFORME

(Core :: Internationalization, defect)

Description

[Henri Sivonen (:hsivonen)]

Build ID: 2002083017

Reproducible: Always

Steps to reproduce:
1) Load the the test case (upcoming)

Actual results:
There are glyphs for all the tested characters in the version of Times the ships
with OS X but not in the version of Times New Roman that ships with OS X. When
Times New Roman is the primary font suggestion, Lucida Grande is used. When
Times is the primary font suggestion, Times is used.

Expected results:
Expected Times to be used in both cases.

Comment 1

[Henri Sivonen (:hsivonen)]

Attachment: Test case

Comment 2

[Niklas Dougherty]

Attachment: Comparison with TextEdit

Actually, this is standard ATSUI behavior, to a certain degree.

Copy the text in the previous attachment and paste it into a Rich text-enabled
TextEdit document. It will probably default to Lucida Grande, and all is fine.
Select Times in the font panel, and all is still fine. However, when selecting
Times New Roman, TextEdit reverts back to Lucida Grande for the characters that
the M$ font lacks.

Mozilla complicates things further. The capital G with caron is not Lucida
Grande, but Helvetica or another, narrower font.

With or without ATSUI, Mozilla must traverse the document defined fonts on a
character to character basis and see to it that the first choice applies if it
is possible, then the next, and so on, and in the end it should default to
whatever ATSUI defaults to if no defined font suffices.

Comment 3

[Henri Sivonen (:hsivonen)]

> Actually, this is standard ATSUI behavior, to a certain degree.

When the app doesn't provide ATSUI with the required information, yes. But the
styles in the test case provide that info, so not passing the info to ATSUI is a
bug.

> With or without ATSUI, Mozilla must traverse the document defined fonts on 
> a character to character basis and see to it that the first choice applies 
> if it is possible, then the next, and so on, and in the end it should 
> default to whatever ATSUI defaults to if no defined font suffices.

Mozilla's job would be to provide ATSUI with the fallback list from CSS. Then it
would be ATSUI's job to do the fallback efficiently.

See
http://developer.apple.com/techpubs/macosx/Carbon/text/ATSUI/Apple_Type_S_code_Imaging/Functions/Mapping_Font_Fallbacks.html#//apple_ref/c/func/ATSUSetFontFallbacks

Comment 4

[Greg K.]

Was that screenshot taken in 1.1? Post 1.1 builds with the fix for bug 111728
shouldn't have that extra padding around the l+slash. I certainly don't see it
using FizzillaCFM/2002082909.

Comment 5

[Henri Sivonen (:hsivonen)]

Attachment: Screenshot from build 2002083017 on Jaguar

Screenshot from a recent nightly.

Comment 6

[Roy Yokoyama]

ftang or nhotta should own this.
nhotta for now.

Comment 7

[nhottanscp]

Reassign to ftang.

Comment 8

[Frank Tang]

Sorry, I am confused, is this problem fixed or not ?
What is the problem in 
http://bugzilla.mozilla.org/attachment.cgi?id=97403&action=view
?

Comment 9

[Henri Sivonen (:hsivonen)]

No, this hasn't been fixed. Both lines in the test case should be rendered in
Times. However, the first line is rendered in Lucida Grande.

Comment 10

[Frank Tang]

assign

Comment 11

[Frank Tang]

what a hack. I have not touch mozilla code for 2 years. I didn't read these bugs
for 2 years. And they are still there. Just close them as won't fix to clean up.

Comment 12

[Travis Chase]

Mass Reassign Please excuse the spam

Comment 13

[Travis Chase]

Mass Re-opening Bugs Frank Tang Closed on Wensday March 02 for no reason, all
the spam is his fault feel free to tar and feather him

Comment 14

[Travis Chase]

Reassigning Franks old bugs to Jungshik Shin for triage - Sorry for spam

Comment 15

[Simon Montagu :smontagu]

I see both lines in the testcase as Times in a current trunk build.