[Webkit-unassigned] [Bug 13709] Table border doesn't show up
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Mon Apr 1 11:49:51 PDT 2013
https://bugs.webkit.org/show_bug.cgi?id=13709
--- Comment #24 from Suchit Agrawal <a.suchit at samsung.com> 2013-04-01 11:48:02 PST ---
(In reply to comment #23)
> (From update of attachment 195988 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=195988&action=review
>
> > Source/WebCore/ChangeLog:10
> > + Border width is currently of type unsigned. The pixel value of 0.7pt or less
>
> This was a decision made when switching to sub-pixel layout to keep the value unsigned (see http://trac.webkit.org/wiki/LayoutUnit). IIRC it was to ensure that borders were always of the author size.
>
> > Source/WebCore/ChangeLog:22
> > + Any result which is less than 1, set to 1 in case of integer templates.
> > + Any result which is less than 1, returnedas it is in case of float templates.
>
> I don't really understand these rules.
>
When the computed value of border width in pixels from pt units is less than 1px, border width becomes 0 due to rounding off double to unsigned. However the border width should not be set to zero unless explicitly specified by the author.
Also for cases where ApplyPropertyComputeLength ::applyValue() has to compute a floating type value, the computed result should not be rounded off as in the above case.
> What are other browsers doing? Are we matching them?
>
> Nothing is said in CSS 2.1 (or CSS 3 borders) about it, I think we should seek out clarification in this case or else we will not reach inter-operable behavior?
>
Varified the behavior on FF and IE8. Webkit matches both FF and IE with this patch applied.
> > Source/WebCore/css/StyleBuilder.cpp:649
> > + }
>
> Do we really need all this copied and pasted code? (LayoutUnit does some similar operations but the core function is kept around).
>
I will check and get back to you on this.
--
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
More information about the webkit-unassigned
mailing list