W3C home > Mailing lists > Public > www-style@w3.org > June 2014

Re: Why not max/min-font-size? & extend them to other properties of sizes

From: Felix Miata <mrmazda@earthlink.net>
Date: Sat, 28 Jun 2014 02:49:03 -0400
Message-ID: <53AE655F.8090009@earthlink.net>
To: www-style@w3.org
On 2014-06-26 23:33 (GMT-0700) Bruno Racineux composed:

> The Normal/Medium size is always 16px by default for all browsers.

Wrong:
http://fm.no-ip.com/SS/brunosFontDefaults1406-144Wie6.png

Wrong:
http://fm.no-ip.com/SS/brunosFontDefaults1406-144L1.png

Wrong:
http://fm.no-ip.com/SS/brunosFontDefaults1406-180L2.png

Prior to the web, fonts were sized almost universally in points, a physical 
unit of measure. They still are, in print media as we all know, but also, as 
the images demonstrate, on computer desktops, and in word processing apps 
used with computers.

Some browsers do indeed size their defaults directly in px, but others work 
out to 16px only by the happenstance that the environmental density either is 
or is assumed to be 96 DPI. These others size like the rest of the world, in 
physical points, which may or may not be accurate, but sized thus nevertheless.

As points are a physical unit everywhere except in web standards, and in 
browsers arguably compliant with them, their physical size in px depends on 
actual physical display px density, which in turn means they won't be 16 
angular px unless they're sync'd up to that magic number 96 (or possibly an 
integer multiple thereof).

The images show, if appropriately viewed in an image viewer adjusting them to 
a physical size matching the original environment, that e.g. in a 144 DPI 
environment the 12pt default size some browsers used to and others do still 
ship with works out to 24px, and at 180 DPI to 30px, where px is defined as 
whatever size the browser may choose call it, if any unit all is presented, 
most often closer to a device pixel than an angular pixel.

The are two basic problems with the widespread assumption that all browsers 
ship with a 16px default:

1-It isn't and never has been[1] true.

2-Making any assumption at all about either the shipped with size or the size 
in fact disregards the issue that making any assumption at all is the wrong 
thing to do, with one very major exception, which is that the default size is 
presumptively optimal reading size for the person using the browser. If in 
fact the shipped default is inappropriate, the user is the only one in 
position to ascertain an appropriate personalization.

[1] e.g. charted how they varied in IE6. It's old, so you need browser and DE 
that adhere to the concept of pt as physical units for the samples to show 
what they showed when created: 
http://fm.no-ip.com/Auth/IE/absolute-sizes-IE6.html Netscape 4 and its 
predecessors, like IE, sized in pt, so their 12pt as shipped defaults were 
only 16px in an actual or assumed 96 DPI environment.

As to $SUBJECT, CSS has already given web stylists an overabundance of power 
to make life difficult for web users. Limiting the ability of users and their 
browsers to override disrespectful web styling and maintain some semblance of 
legibility has no business being any part of web standards.

On 2014-06-17 12:59 (GMT-0700) Tab Atkins Jr. composed:

> 16px is always (approximately) the same size; it's an absolute unit.

As an angular measurement, true.

> It does not depend on the pixel density of your device.

It doesn't, but pixel density is a part of the real world that the 
specification ignores. Its use by stylists brings with it assumptions that 
don't play nicely with usability, accessibility or respect for web users. 
There's wide variation in proximity between the angular specification pixel 
and the real world CSS pixel due to the granularity dictated by browser 
engine and computing power limitations.

Web users would be much happier people if the CSS px unit didn't exist, and 
stylists were forced to use units already at their disposal that are a 
function of size determined by each individual web user (e.g. em/ex/rem, or 
fractions or multiples thereof), leaving it up the the browser engines to 
determine automatically how many pixels to use to achieve them, leaving no 
need for users to reach for size controls on every first page load of a new 
domain, or worse.

> If it did, all pages would have been broken for years already, since devices
> have DPIs ranging from 72 to 300+.

The web has had an overabundance of broken pages for well over a decade, 
nearly as long as CSS has existed, perpetuated by specifications usurping 
ancient names applicable to physical units of measurement to apply to 
illogical units of measure.
-- 
"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 Saturday, 28 June 2014 06:49:27 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:39:22 UTC