Bug 156840

Summary: More improvements and explanations regarding resetting CSS properties on the <math> element
Product: WebKit Reporter: Frédéric Wang (:fredw) <fred.wang>
Component: MathMLAssignee: Frédéric Wang (:fredw) <fred.wang>
Status: RESOLVED FIXED    
Severity: Normal CC: alex, darin
Priority: P2    
Version: WebKit Nightly Build   
Hardware: All   
OS: All   
Bug Depends on: 133603    
Bug Blocks: 156837, 156839    
Attachments:
Description Flags
WIP Patch (for EWS testing)
none
Patch darin: review+

Frédéric Wang (:fredw)
Reported 2016-04-21 02:22:20 PDT
See bug comment 18. This will address: (In reply to comment #18) > 1) Indenting. Will be done here. > > 2) Continuing to use a different list on iOS than on every other platform ifdef will be removed. > 3) Removal of many of the font names that were already listed > 4) Is it really OK to set the font-family without taking font size into > consideration. > I don’t think the comment about the wish to have more built-in fonts on OS X > and iOS adds much here.> > Is the change in behavior on OS X really carefully considered? Seems that it > will now use "serif" and before would have used "Lucida Grande" or maybe > "Arial Unicode MS"? I already explained these in bug 133603 comment 19, but I'll update the comment in the CSS refer to http://trac.webkit.org/wiki/MathML/Fonts explaining why fonts are necessary, problems with obsolete fonts and importance of customization for consistent rendering (not just the font-size). > 5) Overriding the font-family but still inheriting the many other > font-related properties. I'll add a link to bug 156839. Additionally, I think I can use @font-face and Ahem to provide a pixel tests checking math font selection.
Attachments
WIP Patch (for EWS testing) (5.79 KB, patch)
2016-04-21 04:17 PDT, Frédéric Wang (:fredw) 		no flags
DetailsFormatted DiffDiff
Patch (8.00 KB, patch)
2016-04-21 05:07 PDT, Frédéric Wang (:fredw) 		darin: review+
DetailsFormatted DiffDiff
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.
Frédéric Wang (:fredw)
Comment 1 2016-04-21 04:17:34 PDT
Created attachment 276907 [details] WIP Patch (for EWS testing)
Frédéric Wang (:fredw)
Comment 2 2016-04-21 05:07:00 PDT
Created attachment 276910 [details] Patch
Darin Adler
Comment 3 2016-04-21 18:52:26 PDT
Comment on attachment 276910 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=276910&action=review Seems OK. It seems really peculiar to list fonts that are almost all available together. Given the discussion here there is clearly a call, longer term, for some clever system to choose which font to use based on which non-math font is being used in the surrounding content. > Source/WebCore/css/mathml.css:73 > + We put Cambria Math first since it is pre-installed on Windows 7 and higher. Not sure I understand the logic here. No reason I can think of to put more-often-installed fonts first.
Frédéric Wang (:fredw)
Comment 4 2016-04-21 22:48:54 PDT
(In reply to comment #3) > It seems really peculiar to list fonts that are almost all available > together. Given the discussion here there is clearly a call, longer term, > for some clever system to choose which font to use based on which non-math > font is being used in the surrounding content. Definitely. For now, the idea is to not miss any math font that can be installed on the system. > > Source/WebCore/css/mathml.css:73 > > + We put Cambria Math first since it is pre-installed on Windows 7 and higher. > > Not sure I understand the logic here. No reason I can think of to put > more-often-installed fonts first. As I said on bug 133603, this is a bit arbitrary... Maybe a better reason is because Cambria Math is pre-installed it has been more tested than the two others.
Frédéric Wang (:fredw)
Comment 5 2016-04-21 22:56:28 PDT
Committed r199869: <http://trac.webkit.org/changeset/199869>
Frédéric Wang (:fredw)
Comment 6 2016-04-24 07:34:33 PDT
(In reply to comment #3) > It seems really peculiar to list fonts that are almost all available > together. Given the discussion here there is clearly a call, longer term, > for some clever system to choose which font to use based on which non-math > font is being used in the surrounding content. For the record and future discussions, I've opened an issue in the MathML in HTML5 implementation note: https://github.com/MathML/MathMLinHTML5/issues/26 Ideally, I'd like CSS properties/keywords that could just be used in the default user agent stylesheet to tell the web engine to perform such clever trick. That way users won't be surprised by a non-CSS compatible behavior or can at least override that behavior if they don't like it.
Note You need to log in before you can comment on or make changes to this bug.