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

RE: [css-display-viewport] Enabling fast layout for web apps via isolation

From: François REMY <francois.remy.dev@outlook.com>
Date: Thu, 20 Dec 2012 20:32:49 +0000
Message-ID: <DUB404-EAS320F90B9D3CBC4C72762D1EA5370@phx.gbl>
To: Ian Hickson <ian@hixie.ch>
CC: CSS WG <www-style@w3.org>, Tobie Langel <tobie@fb.com>, Tab Atkins Jr. <jackalmage@gmail.com>


Actually, my proposal looks like an “overflow: hidden; position: relative;” element and where ‘position: fixed’ is remapped as “position: absolute(from-local-viewport)”.

If you don’t use  “position: relative” on the ‘local viewport’, an absolutely positioned element inside the viewport depends on the latest “position: relative” element.

If you don’t remap “postion: fixed” you’re screwed again. This last part is the essential one that you can’t achieve in CSS right now, and that’s why you need a new layout engine.


In fact, the real innovation is that, for the first time, you can be sure without any heuristics or checking that the element can be considered as a replaced element by the outer layout manager.



I’m not quite sure why we should restrict only to ‘transform: translate()’ positioning inside a viewport. I can easily understand that this is a quick and dirty way to achieve hardware-accelerated rendering in Webkit, but to me this seems like an artificial restriction imposed by the webkit engine, not really one that we should get in the spec. I would be interested in the opinion of other implementers there. 

For example, I think that a grid layout may be very appropriate as well (and you can perfectly add new rows to the grid without having to relayout the previous rows, and only load in memory the visible rows as well as some of the next and previous ones).

I’m not saying JavaScript-based layout managers are bad (actually, I think they’re going to be more popular over time, particularly when tools will support them better) but that most designers will probably want to use the layout managers they know and love instead of relying on JavaScript code they can’t understand.




De : Ian Hickson
Envoyé : ‎20‎ ‎décembre‎ ‎2012 ‎18‎:‎35
À : François REMY
Cc : CSS WG, Tobie Langel, Tab Atkins
Objet : Re: [css-display-viewport] Enabling fast layout for web apps via isolation

On Thu, 20 Dec 2012, François REMY wrote:
>        http://fremycompany.com/TR/2012/ED-css-viewport/

How does this differ from overflow:hidden? (Not saying it doesn't, just 
curious as to what the difference is.)

Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Thursday, 20 December 2012 20:33:17 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:08:24 UTC