W3C home > Mailing lists > Public > www-style@w3.org > October 2011

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

From: François REMY <fremycompany_pub@yahoo.fr>
Date: Fri, 28 Oct 2011 19:07:40 +0200
Message-ID: <20D9ACC01E7C4ACD8AB4EA0750BAB7EE@FREMYD2>
To: "Boris Zbarsky" <bzbarsky@MIT.EDU>
Cc: "CSS 3 W3C Group" <www-style@w3.org>
± -----Message d'origine----- 
± From: Boris Zbarsky
±
± On 10/28/11 4:07 AM, François REMY wrote:
±  > That means that a CSS pixel could ends up to be a non-round
±  > number of physical pixels. But if it's what the developer asked.
±
± Chances are, developers would just ask this as boilerplate because it
± works fine (i.e. ends up as integer physical pixels) on their machine.
± And then for people who actually end up with noninteger pixels the site
± will be very slow and/or blurry.
±
± As a UA developer, I would be very unlikely to implement something which
± would almost universally be used to screw over my users like that.

Indeed. I didn't think about that one.

± In any case, as pixel densities increase this should become more and
± more of a nonissue.
±
±  > For the second part, maybe I wasn't as clear as I intended. I would
±  > introduce a "new kind" of DPI that is defined as "equivalent dot per
± inch if
±  > on a screen locate 0.5 meter away".
±
± That's more or less the old definition of CSS px; the one that was
± changed because it caused the problems listed above.

I don't get it. If I can get a "equivalent standard-dpi size" of my device,
I'll certainly be able to make the good choices about my web layout.

I doesn't mean the the pixel definition *HAS* to be related to that 
perceived
dpi (they may relate to something different if it makes sense) but at least
I can deal with it.

For exemple, if you have a touchwall device like Microsoft Surface that's
designed to be used at a short distance and another one which sports the
same resolution and pixel definition but that's used for pitchs, it's not a
big issue for me... if I can get a clue about that.

For both device, I will probably get a 96dpi resolution, but I want to be 
able
to get a "equivalent standard-dpi". For the Surface one, I would probably 
get
an equivalent of 96 eDPI because people usually stand up at 0.5 meter of the
screen. For the slideshow display, however, I would probably get 170 eDPI
because poeple are located much further.

Then, you may specify to the UA what you really want to do with your CSS
pixel. Get a true "1 physical pixel = 1 css pixel" (I build my stuff for one 
device),
get a "1 css pixel ~ 1/96 inch of the screen", or get a "1 css pixel ~ the 
best
unit to get roughly the same perceived size as a css pixel on a screen which 
is
located at 0.5 meter of the reader".

François 
Received on Friday, 28 October 2011 17:08:31 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:45 GMT