RE: [css3-floats] Comments on floats/positioning draft

It can be 2-pass only. The first time around you "collect" the positioned exclusions and their static positions and the second time you layout with the exclusions present. 

You can do even better in cases when the positioned elements don't rely on their static positions or you know that there aren't any other exclusions above a given node (in this case you can restart that sub-tree only).

As you pointed out however, if you want to do everything as you go, you will get exponential complexity for the most inner nodes.

Thanks,
Rossen Atanassov

>-----Original Message-----
>From: www-style-request@w3.org [mailto:www-style-request@w3.org] On
>Behalf Of Boris Zbarsky
>Sent: Wednesday, August 03, 2011 1:29 PM
>To: www-style@w3.org
>Subject: Re: [css3-floats] Comments on floats/positioning draft
>
>On 8/3/11 4:24 PM, Rossen Atanassov wrote:
>> In short. The processing model as of now is a 2-pass layout exactly as you
>outlined it below.
>
>Any time I see that sort of thing, I get worried.  In particular, every instance of
>actual two-pass layout in Gecko that was web-facing has had to be more or less
>removed because of sites nesting things needing such layout, leading to
>exponential growth in the number of passes on the innermost boxes.
>
>So is this the sort of two-pass layout that's only two passes no matter what the
>nesting depth is, or is this the sort that gives you 1024 passes once nested 10
>deep?
>
>-Boris
>

Received on Wednesday, 3 August 2011 20:42:13 UTC