W3C home > Mailing lists > Public > www-style@w3.org > February 2012

Re: [css3-values] Physical length units

From: Charles Pritchard <chuck@jumis.com>
Date: Sun, 19 Feb 2012 12:47:23 -0800
Message-ID: <4F415FDB.6060605@jumis.com>
To: Lea Verou <leaverou@gmail.com>
CC: Felix Miata <mrmazda@earthlink.net>, www-style@w3.org
On 2/19/12 12:37 PM, Lea Verou wrote:
> On 19/2/12 22:18, Charles Pritchard wrote:
>> On 2/19/12 11:42 AM, Felix Miata wrote:
>>> On 2012/02/19 09:15 (GMT-0800) Tab Atkins Jr. composed:
>>>> It's largely people using the pt unit for fonts, but the other
>>>> units show up at times, and people expect them to have a dependable
>>>> ratio with px.
>>> Others expect measuring units defined centuries ago to retain the 
>>> same meaning in novel contexts as they do in traditional contexts. 
>>> cf. the puter world's hijacking of decimal multiples MB, KB, GB, 
>>> etc. http://en.wikipedia.org/wiki/Binary_prefix
>> Is there any pushback with the truemm concept? I've seen this thread 
>> pop up for awhile [years?] now.
>> RoC put out a definition, it lacks any handling for browser zoom (if 
>> the browser is zoomed in 2x, is it still 1mm?):
>> http://lists.w3.org/Archives/Public/www-style/2010Jan/0343.html
> Of course not. Zooming is zooming, so I'd expect it to be 2mm with a 
> 2x zoom.

I want to make sure that part is firmly stated in the specs.

>> I got a lot of pushback from Mozilla when I asked for pixel ratio to 
>> be exposed in 2010/2011. Their entire model, up to the point they 
>> introduced media queries, is built around not exposing this kind of 
>> information. Once they introduced media queries, it simply became 
>> obfuscated instead of obstructed. Mozilla was the only browser to 
>> take such a rigid position, and it has loosened.
> Since right now screen dpi can easily be detected by a simple binary 
> search with resolution media queries in Gecko, I guess they're fine 
> with it now. :)

It's no fun though. About 15 lines of CSS.

>> Are there any other use cases than the simple "put a ruler on the 
>> screen" demo? I've used that one, cause I've worked with drawing 
>> apps. If that is the only use case -- is simply exposing this 
>> information in window.screen sufficient? I'm fine with css units too, 
>> but my uses are going to be via scripting.
> Any kind of length, really! Font-sizes, widths, heights, anything we 
> want to have the same size in all screens (and we usually do). I 
> remember this kind of size difference across screens was one of my 
> first bad surprises when starting in this field. I think when support 
> for such a unit becomes reliable across browsers/versions, we will 
> start seeing it being used for everything, kinda like what happened 
> with the em unit a few years ago after people realized how it could be 
> used.

For the most part, I'd prefer authors try to stick with "em". We've got 
a lot of flexibility we're trying to get into design. Many big players 
fail to test their site with browser zoom. I don't see why someone needs 
to say, well I want this letter to be 8mm tall; other than using 
relative ratios. And "em" can be good for that. It's a use case, but 
it's not a strong one. It is already poorly executed with "px".

Font sizes is one place where users frequently override what's on the 
page. If we had truemm, I'd add a note to WCAG-TECHS about its 
accessibility. Something like, truemm should only be used for decorative 
content when used with font size.

It's really frustrating to use a site where fine motor movements are 
required for basic navigation. I'd prefer it if everything on the web 
were 1.5x larger and more legible. Thank goodness for mobile phones. At 
least they have people considering that sites should be easy to use with 
big targets to tap on.
Received on Sunday, 19 February 2012 20:47:47 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:08:12 UTC