- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Thu, 11 Aug 2011 08:43:45 -0700
- To: Alex Mogilevsky <alexmog@microsoft.com>
- Cc: Øyvind Stenhaug <oyvinds@opera.com>, "www-style@w3.org" <www-style@w3.org>
On Thu, Aug 11, 2011 at 6:47 AM, Alex Mogilevsky <alexmog@microsoft.com> wrote: > As for "position:relative", in this hypothetical situation (of defining display and positioning properties as if they didn't exist yet) I would keep it separate, as it isn't really specifying a position, it is an offset from a position that is determined before it applies. > > Perhaps something like this would work better for relative positioning: > > position-offset: <length> <length>? > > (the two values are offsets for 'start' and 'before'; somewhat similar to http://www.w3.org/TR/css3-gcpm/#the-float-offset-property) I agree with François - position:relative was a way to do a purely-visual translation of an element before transforms existed. Now that they do exist, if we were doing it over again, we'd just use transform. (We'd also need to separate out the creation of positioning containers, as that's the other useful effect of relpos, and in fact the major reason I use the value. That's a pretty easy-to-justify property, though.) ~TJ
Received on Thursday, 11 August 2011 15:44:33 UTC