Re: Link style sheets

Craig Hubley (craig@passport.ca)
Sun, 14 May 95 18:26:41 EDT

>> If you want the correspondence between 'relationship' and 'function' to
>> live in a style sheet, fine, as long as there is a place for it there.
>> Is there ?
>
> Your original proposal was to invent a new method instead of using http.

Yes, because HTTP doesn't allow for the browser to 'intercept' and
'interpret' the name without going to the network.

> As with style sheets, that's unneeded. The link style sheet is, yes,
> the place where "the correspondence between 'relationship' and
> 'function'" should live. As for whether there's a place for it,
> you'll have to invent the link style sheet first. Perhaps DSSSL Light

I was trying hard not to 'invent' anything.

> has some possibilities, perhaps you need to do something else.
> Don't make us solve the link style sheet format issue in HTML.

Seems to me we have a "mine shaft gap"... :-)

> There is no reason not to arrive at common semantics for such things
> as LINK REL=NEXT, and every reason to try to do so.

No objection to common semantics, we should start with the list of functions
that the browsers already impose on navigation: back, home, etc., so authors
can 'tell' the browser what these mean in the context of a particular document.
That's what SCO has done.

>Application-specific info doesn't need to be standardized in the HTML DTD.

No, but any generic mechanism for resolving it, probably has to be referenced
in it. Link types are supposed to be an RFC *outside* the HTML DTD, anyway,
the only reason we continue this discussion in html-wg is that we were asked
to do so, to keep everyone informed.

-- 
Craig Hubley                Business that runs on knowledge
Craig Hubley & Associates   needs software that runs on the net
mailto:craig@hubley.com     416-778-6136    416-778-1965 FAX
Seventy Eaton Avenue, Toronto, Ontario, Canada M4J 2Z5