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

On 10/5/11 10:17 AM, Brian Blakely wrote:
> Unless these devices are all pixel-doubling and
> -quadrupling without exception

If a CSS pixel on those devices is big enough they should be, no?

For the particular case you link to, it's a 316dpi screen meant to be 
held at reading distance (slightly bent arms, etc).  15in is a good 
estimate for the distance, I believe, but correct me if I'm wrong.

Computer screens are normally a bit further away than that, typically; 
figure 25 inches.

So a CSS px (which is an angle subtended by 1/96 in at computer screen 
distance) is about (1/96 in)/(25in).

A device pixel on this device subtends an angle of (1/316 in)/(15in).

So a CSS pixel is 316*15/(96*25) = 1.975 device pixels on this device, 
and should probably be rendered as two device pixels.

> Anyone from representative groups (Mozilla, Chrome Team, Apple) want to
> comment?  Does the spec need to be changed?

I assume you've read the thread starting at 
http://lists.w3.org/Archives/Public/www-style/2010Jan/0058.html and 
especially 
http://lists.w3.org/Archives/Public/www-style/2010Jan/0343.html ? I 
believe that last is more or less Mozilla's official position, and is 
what we aim to implement last I checked (we support "truemm" as "mozmm" 
for now, pending standardization).

-Boris