W3C home > Mailing lists > Public > www-style@w3.org > August 2006

Re: Selector for parent/predecessor?

From: Alastair Campbell <ac@alastc.com>
Date: Mon, 21 Aug 2006 22:47:47 +0100
Message-ID: <44EA2A03.2020307@alastc.com>
To: www-style@w3.org

Mark Birbeck wrote:
> Ah...I misunderstood you. When you said "ample evidence" I thought you
> were referring to studies, reports, statistics...

Not something that is easy to survey, but if you look though:
http://code.google.com/webstats/

- 24% don't specify alt text on images.
- most pages use presentational attributes.
- Table and scripts are more used than divs.

None of which are indications of good practice or trust. (I believe Ian 
conducted this survey, which could explain his dis-enchantment with code 
on the web ;)

> you can't be sure that authors cannot be
> trusted with features that are denied them.

As an author, I would want to produce an effect and look-up how to do 
it. If it worked (then and there), I would use it and move on. If later 
on it turned out to cause a performance headache, I wouldn't be looking 
at the CSS to start with. (Display bugs, yes. Rendering performance, no.)

In an environment where (other) people can build something, test it 
locally in one browser, and be happy that it works for them, I am 
happier with a conservative approach by browser makers.

Not a
Received on Monday, 21 August 2006 21:48:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:46 GMT