- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Thu, 2 Oct 2014 16:42:41 -0400
- To: Simon Pieters <simonp@opera.com>
- Cc: www-style list <www-style@w3.org>
On Wed, Sep 24, 2014 at 3:57 AM, Simon Pieters <simonp@opera.com> wrote: > On Tue, 23 Sep 2014 17:24:35 +0200, Tab Atkins Jr. <jackalmage@gmail.com> > wrote: >> On Tue, Sep 23, 2014 at 3:34 AM, Simon Pieters <simonp@opera.com> wrote: >>> >>> On Tue, 23 Sep 2014 00:55:59 +0200, Tab Atkins Jr. <jackalmage@gmail.com> >>> wrote: >>>> >>>> It sounds like you're missing the definition of the "canvas" itself. >>>> The canvas is defined at >>>> http://www.w3.org/TR/CSS21/intro.html#the-canvas, but it doesn't seem >>>> to quite go into enough detail to nail down that the "origin" of it >>>> would be the rectangle it ordinarily limits rendering to. >>> >>> >>> >>> That's not a correct definition, I think. See the test case above. The >>> origin needs to be the point on the canvas where the top left corner of >>> the >>> viewport is when the viewport is in the default scroll position, or some >>> such, where the default scroll position depends on writing >>> mode/direction. >> >> >> Yes, that's the "finite region of the canvas" that "rendering >> generally occurs witihin". It's not defined anywhere near >> specifically enough for your purposes, but that's definitely the >> concept you need; we just need to put down a better definition of the >> canvas and its "finite region" somewhere so you can point to it. > > > OK. Thanks. Let me know when there is something I can reference. :-) > > How do you explain the element case? Same deal, isn't it? The containing block generated by an element is based on the element's geometry, not its scrollable area. ~TJ
Received on Thursday, 2 October 2014 20:43:31 UTC