W3C home > Mailing lists > Public > site-comments@w3.org > April 2011

Re: Correction, or update needed?

From: Felix Miata <mrmazda@earthlink.net>
Date: Mon, 25 Apr 2011 10:38:15 -0400
Message-ID: <4DB58757.1090105@earthlink.net>
To: site-comments@w3.org
On 2011/04/24 16:47 (GMT) Linda A. Walsh composed:

> One of the pages that came up was:
> http://www.w3.org/QA/Tips/font-size

> By my understanding, there is a minor, but important "bug" on this page that should be corrected.

> About mid-page, there is a section "Units: avoid absolute length units for screen display".

> Under that the first sentence of the first bullet point says:

> "Do not specify the font-size in pt, or other absolute length units for screen stylesheets."

> This statement concerns me since people might think they should substitute 'px', instead, which, IMO, is worse, and 'pt' is more easily resizable as it's not *really* (in practice), absolute.

> If they use 'pt', I can  compensate by setting my "DPI" value,  and in that sense, "pt" is really a relative unit, as it is relative to the "DPI" value.  Theoretically, it should be a non-issue if the DPI was always set correctly for a given monitor,  however, the 'pt' really equal to 1/72 of the value in "DPI".  I.e. if DPI is set to 72, then 1 pt=1 dot (or pixel), but if DPI is set to 144, then the PT is equal to two dots.

> This is especially important "today", since to resize one's display on Win7, MS resets the DPI value to a higher number -- meaning it's not really the number of dots/inch, but a scaled value.

> Of course in HTML5, even 'px' is scaled, but that's even farther in the future (as far as being widely adopted as a standard).

> One could argue that since 'px' is no longer based on the hardware size of a pixel (in HTML5), that maybe 'pt' should also be redefined toward some more useful definition, like 1/12th or 1/16th of 1EM?

It's already been redefined by the latest versions of major web browsers: 
http://weblogs.mozillazine.org/roc/archives/2010/08/css_units_chang.html

> Anyway, the advice as written stuck out to me as 'wrong' (even though, by definition it is not) since in practice the DPI value is now used as a way to rescale not just webpages, but entire displays and most certainly, I've always preferred sites that use 'pt' vs. 'px', since the 'pt' value honors my DPI setting, whereas sites that use 'px', often appear too small (especially when I was using a laptop with a 144DPI screen, *ouch*!)  Now, my monitor is nominally 100DPI, but most text still appears too small, so I generally use a 120DPI value.

The whole situation now is a mess. The relativity assigned by CSS to the px 
unit was relative to an object the designer has no knowledge of, much less 
control over. Even users don't have more than token control, as they must 
acquire a different physical display to change the object to which it is 
relative. To call the px unit relative is wholly orthagonal to the relativity 
recommended for use with CSS styling, and counter productive to achieving a 
better web experience for the widest possible expanse of web users.

Now that the pt to px ratio is fixed at a 3/4 by web browsers, both px and 
pt, as well as cm, in, mm and more, are functional absolutes without regard 
to anything that matters to an end user, particularly no longer WRT DPI, and 
certainly without regard to the size of pt, in, cm, etc. on a printed page.

px as a unit for sizing of more than a single digit should be deprecated in 
CSS3, and removed from CSS4.
-- 
"The wise are known for their understanding, and pleasant
words are persuasive." Proverbs 16:21 (New Living Translation)

  Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/
Received on Monday, 25 April 2011 14:38:37 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 24 October 2012 16:21:33 GMT