RE: LINK'ed style sheets
> From: Gavin Nicol <firstname.lastname@example.org>
> Date: Tue, 18 Jun 1996 05:56:26 GMT
> Cc: email@example.com, firstname.lastname@example.org, email@example.com,
> >Are there alternate solutions to this problem that work with the loads
> >of existing content out there now? Thats the problem with growing a
> >spec organically, there's all that existing-use. its not a W3C problem,
> >its an Internet problem.
> As for alternatives, I see two:
> 2) Use a single *empty* STYLE tag, and put the style information
> into attributes.
This seems like a good solution to me. You could also use a #CONREF
attribute to allow this in addition to the current method. (For those
not familiar with CONREF, when you declare an attribute as #CONREF,
then any element for which you specify this attribute is treated as
having a declared content of EMPTY.) This would mean that as more
browsers supported STYLE elements, people could switch over from
putting it in an attribute to putting it in content.
Both SGML and CSS allow literals to be delimited with either single or
double quotes, and SGML character references can be used to handle any
remaining delimiter conflicts.
I'm sure the ERB must have considered this alternative. What was the
objection to it?