In article <9uqjtt$[EMAIL PROTECTED]>, Mike Stockman 
<[EMAIL PROTECTED]> wrote:

> 2) The bug you refer to suggests that the style sheet should be ignored 
> (if the server's mime type is wrong) only if the document spec is 
> "strict," which mine isn't. Am I misunderstanding?

You are misunderstanding. The bug is about the standards mode. The XHTML 
1.0 Transitional doctype activates the standards mode.

> 3) This document is XHTML Transitional, and the style sheet is being 
> ignored. I have HTML 4 documents on the same server whose style sheet is 
> being used. This seems inconsistent. Is it?

Inconsistent, sure, but for a reason: avoiding the issue know as bug 
22274. (Unless something changes wrt. bug 22274, any change to the 
doctype sniffing semantics would make things worse from some point of 
view. This has been discussed over and over again. Please, let's let the 
mode selection stay the way it is unless something changes wrt. bug 
22274.)

HTML 4.0 Transitional doctypes active the quirks mode, but HTML 4.01 
Transitional doctypes with a system identifier activate the standards 
mode in 0.9.5 and later. In 0.9.4 and earlier, the system identifier had 
to ba a perticular string.

See http://www.hut.fi/u/hsivonen/doctype for details.

> If this style sheet issue works as expected (although possibly against 
> spec) in 0.96, but is broken (or fixed re: spec) in recent builds, is 
> that a problem or a solution?

Good question.

I'm generally all for strict standards compliance and honoring the 
Content-Type header. However, I'm afraid that some people are going to 
use the effect of bug 46225 as an excuse to make even harder to activate 
the standards mode. That would be bad. Then authors might give up with 
Mozilla's standards mode and focus on Mac IE 5, Windows IE 6 and Opera 5 
& 6, which would make Mozilla appear to be to quirky browser.

-- 
Henri Sivonen
[EMAIL PROTECTED]
http://www.hut.fi/u/hsivonen/

Reply via email to