Re: User Agents Do Not Implement Absolute Length Units, Places Responsive Design in Jeopardy

The problem is that you may want to rely on the "device size" to know if you 
need to show a "Mobile optimized" website, a "Tablet optimized website" or a 
"plain old website". That depends on the phyical size of the device, not its 
number of pixels. You can get a tablet of 7inches having as much pixels as 
an iPhone 5 in landscape mode. But the two clearly don't deserve the same 
website. This is a huge problem.



-----Message d'origine----- 
From: Tab Atkins Jr.
Sent: Wednesday, October 05, 2011 5:46 PM
To: Brian Blakely
Cc: www-style@w3.org
Subject: Re: User Agents Do Not Implement Absolute Length Units, Places 
Responsive Design in Jeopardy

On Wed, Oct 5, 2011 at 7:17 AM, Brian Blakely <anewpage.media@gmail.com> 
wrote:
> See spec for a refresh: http://www.w3.org/TR/css3-values/#absolute-lengths
> As far as I know, UAs have never actually implemented this, but always
> pretended to anyway.  If you size something as "1in", you're more than
> likely going to get 90px, regardless of the accuracy of this output.

No, you're going to get exactly 96px.  Several (all?) of the UAs agree
on this now.


> This is important because physical screen dimensions are going to become
> very crucial in the near-term for content and layout delivery, especially 
> in
> regards to Media Queries.
> The problem which was years away is now an impending reality: the day when
> 720p+ mobile devices become the norm.  Unless these devices are all
> pixel-doubling and -quadrupling without exception, using Media Queries for
> the most common responsive design practice — horizontal resolution 
> detection
> in pixels — won't work anymore.  They will simply cease to be a solution,
> leaving no replacement.

Once we reach 720 dpi or so, the problem pretty much solves itself
(actually, that should happen somewhere around ubiquitous 300dpi).  At
that point, the maximum difference that "1in" can be from a physical
inch is pretty small.  Resolution-detection using pixels stops
working, but that was always an accidental benefit.


> We need a new way to do generic device detection, or we need absolute 
> length
> units to work.
> Anyone from representative groups (Mozilla, Chrome Team, Apple) want to
> comment?  Does the spec need to be changed?

Can you describe the problem more directly?  Do you need better ways
to do resolution detection?

~TJ 

Received on Wednesday, 5 October 2011 16:04:31 UTC