W3C home > Mailing lists > Public > www-style@w3.org > September 2012

Re: Used Style Computation (and Viewport/Canvas Layout)

From: Boris Zbarsky <bzbarsky@MIT.EDU>
Date: Tue, 25 Sep 2012 08:32:02 -0400
Message-ID: <5061A442.5050104@mit.edu>
To: François REMY <fremycompany_pub@yahoo.fr>
CC: "www-style@w3.org" <www-style@w3.org>
On 9/25/12 3:55 AM, François REMY wrote:
> Seems like an "overflow: auto" issue to me, not an "overflow: visible".

Well, the point is that given the spec for overflow:auto, if you have an 
element that you don't want to leak layout stuff to ancestors you have 
to make sure its overflow is not "visible".

> And it only happens on browsers that support scrollbars.

Well, yes, which is obviously an edge case and everything, sure.

> Also, what if the "display: viewport" element would be considered to consist in two Layout Tree nodes? One for the "outside" layout which only takes its raw width/height/border/margin/padding in consideration and another one for the "inside" layout which is its (only) direct child and which deals with the children exclusively?

How does that change anything?  You would still need to effectively give 
non-visible overflow to the inner node, right?  (Incidentally, what you 
just described is exactly how text inputs, say, work in Gecko.)

-Boris
Received on Tuesday, 25 September 2012 12:34:15 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:21:00 GMT