RE: LINK'ed style sheets

>>>SGML comments are now a valid way to "hide" the <STYLE> contents from
>>>non-conformant browsers, as decided by the W3C.
>>
>> This is, to put it simply, the wrong way to do things. Even PI's
>> are preferrable.
> 
>  Amen hallelujah to that one.  Hacking your parser to not parse comments
>inside a <style> block is just evil.  I'm not an SGML stud, but is there
>_ANY_ way to describ e this behaviour with a DTD fragment?
 
No.

I *am* an "SGML stud", but I've pretty much given up on getting people
to respect even simple things, however the above is a gross hack. Even
Dan's harily little flex scanner needs to have code added on top to
handle this.

I'm very worried about the W3C. They have an aire of respect, so
people listen to them. However, so far they've put out HTML 3.2, to
the surprise and disappointment of many, and CSS despite some people
having serious problems with it, and now they recommend non-conformant
parsing. Perhaps they should stop pretending that they care about SGML
conformance, or at least state publically that the main focus for them
is to make vendors happy.

Received on Monday, 17 June 1996 23:15:47 UTC