Re: [CSS21] Test Suite

On Friday, July 23, 2004, 10:52:37 PM, Henri wrote:

HS> On Jul 23, 2004, at 22:15, Chris Lilley wrote:
>> BZ> What behavior do other browsers have with regard to unknown 
>> doctypes?
>>
>> BZ> This would be a good thing to add to that quirks/standards chart 
>> that was cited
>> BZ> earlier in the thread.
>>
>> It would.

HS> My policy has been to be intentionally silent about old doctypes, 
HS> obscure doctypes and homegrown doctypes in order to discourage people
HS> from using them.

Well, that is fair enough; but for XHTML its certainly appropriate to
use a different one, for example to describe which subset of that
modular specification is used.

HS> I seriously recommend using only either
HS> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" 
HS> "http://www.w3.org/TR/html4/loose.dtd">

Which gives standards mode in some browsers and quirks mode in others.
Which is a pity.

HS> or
HS> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" 
HS> "http://www.w3.org/TR/html4/strict.dtd">
HS> for new content served as text/html.

I tend to use XHTML 1.0 Strict for all text/html content, unless I need
a link to open in another window in which case I have to go to
Transitional.

HS> This recommendation of mine has been stable and appropriate for four
HS> years despite refinements to doctype sniffing and more browsers 
HS> adopting the practice. I see no good reason to encourage authors to use
HS> a wider variety of doctypes for text/html. (At least not until the WHAT
HS> WG work exits the draft stage.)

HS> (For XML on the Web, including XHTML served as application/xhtml+xml, I
HS> advocate doctypelessness and using other means, such as Relax NG, for
HS> assessing correct element usage.)

I agree on both counts.

>> Results from more browsers, if available, would be good there
>> too. I am thinking particularly of Safari,

HS> Already covered in the same column as contemporary Mozilla.

Thanks (although that was not clear when I posted my question)

>> Konqueror,

HS> On my todo list.

Good.

>>  NetFront and the browser from Openwave.

HS> Are these known to do doctype sniffing?

They are not known not to. They are known to do CSS, so seem to be worth
finding out more.





-- 
 Chris Lilley                    mailto:chris@w3.org
 Chair, W3C SVG Working Group
 Member, W3C Technical Architecture Group

Received on Friday, 23 July 2004 17:28:56 UTC