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

Re: The CSS Problem

From: Joshua Cranmer <Pidgeot18@verizon.net>
Date: Mon, 12 Nov 2012 23:56:50 -0600
Message-id: <50A1E122.6070006@verizon.net>
To: www-style@w3.org
On 11/12/2012 8:46 PM, Jens O. Meiert wrote:
>>> <http://meiert.com/en/blog/20121112/the-css-problem/>.
>> You don't suggest an alternative.
> I believe I do, by suggesting a different focus on the Working Group
> side (what is critical, and hence, implied, what can be removed?), and
> also more focus on the community side (how can we get the most out of
> CSS without constantly screaming “more, more”?).

The adage that 80% of your users use only 20% of your features comes to 
mind, as does the equally important adage that they all use a different 
20% of them.

Looking at the list of CSS current-work specifications, I would consider 
backgrounds & borders, multi-column, and flexbox to be the only 
"critical" specifications for my needs. Judging from mailing list 
traffic (i.e., how many times the same feature has been re-requested in 
various forms) over the past several years, parts of Selectors 4, CSS 
conditionals, and variables would rank as "critical" work. Judging from 
code that has found its way on the web, animations, transitions, and 
transforms seems to have fallen into the category of "critical" as well.

I'll also point out that judging complexity from the number of 
properties is a poor proxy, as the marginal difficulty of implementing 
border-left over border-right is far less than the marginal difficulty 
of implementing a variable property, and the extra cost in adding a new 
display: type can be as large as most properties.

Beware of bugs in the above code; I have only proved it correct, not tried it. -- Donald E. Knuth
Received on Tuesday, 13 November 2012 05:57:22 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:08:23 UTC