W3C home > Mailing lists > Public > www-style@w3.org > December 2010

Re: [CSS21] 4.3.2 Lengths (reference pixel?)

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Wed, 15 Dec 2010 08:38:32 -0800
Message-ID: <AANLkTinYe1sbumR=Ns+sjvU1ag4h=0mZzFShWvzLEvBG@mail.gmail.com>
To: Boris Zbarsky <bzbarsky@mit.edu>
Cc: Ambrose LI <ambrose.li@gmail.com>, www-style@w3.org
On Wed, Dec 15, 2010 at 8:24 AM, Boris Zbarsky <bzbarsky@mit.edu> wrote:
> On 12/14/10 11:48 PM, Ambrose LI wrote:
>> If we cannot rely on 1 image pixel = 1px
>
> You can rely on "pixels" in bitmap images being treated as CSS pixels by a
> CSS renderer.  For example, if you take a 100px by 100px JPEG image and
> print it on a 600dpi printer, it's not going to end up 1/6 inch wide and
> tall!  Nothing is changing there.  Bitmap image pixels continue to be
> divorced from device pixels, in general.

Which is, of course, one part of the reason why CSS px are always an
integer multiple of device pixels, which is what's causing the current
kerfuffle.  If they're an integer multiple, then images can be
rendered with trivial scaling; non-integer multiples requires more
complex scaling algorithms that would introduce blur.

There are a lot of good reasons why the current state of the world is
the way it is.  ^_^

~TJ
Received on Wednesday, 15 December 2010 16:39:26 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:35 GMT