- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Fri, 4 Dec 2009 08:16:19 -0600
- To: Boris Zbarsky <bzbarsky@mit.edu>
- Cc: Patrick Garies <pgaries@fastmail.us>, www-style <www-style@w3.org>
On Fri, Dec 4, 2009 at 7:56 AM, Boris Zbarsky <bzbarsky@mit.edu> wrote: > What Gecko implements is the following, more or less: > > If computed direction is ltr, then the computed value of padding-left > is given by the more specific of the specified values of > padding-left and padding-start. > > If computed direction is ltr, then the computed value of padding-right > is given by the more specific of the specified values of > padding-right and padding-end. > > And similar for the rtl case. I haven't tested what webkit implements, but > I suspect it's similar. So basically *-start and *-end are treated as context-dependent aliases for *-left and *-right? That's how I imagined it would work, and that does make the most sense and is likely the simplest way to address the problem. ~TJ
Received on Friday, 4 December 2009 14:16:54 UTC