Re: SGML and HTML

"Jukka K. Korpela" <jkorpela@cs.tut.fi> wrote in message
news:Pine.GSO.4.50.0309041920370.11495-100000@korppi.cs.tut.fi...
>
> Should the "standard" be changed, then? Well, they just did. XHTML makes
> the verbose canonical syntax a must. So it's somewhat strange that we are
> now _also_ witnessing a change where a purported validator is changed to
> process _documents declared as HTML 4.01_ according to "practical rules".

XHTML is utterly impractical to be deployed on the web since so few user
agents support application/xhtml+xml remotely adequately (I'd say Opera
only.) which leaves us with either the ludicrousness of XHTML 1.0 Appendix
C. +browser realities or HTML 4.01 + browser realities.   Since Appendix C.
is gibberish which talks about "most browsers", and gives examples which
would violate text/* mime-type defaults, I don't think we can do much with
that - certainly not until there's a QA tool to help us check we are
conformant to Appendix C at the very least.

So that leaves the only option as HTML 4.01 - We then need to deal with the
browser realities which is such that HTML is not an application of SGML,
hopefully that claim will get Errata'd away or something, as you say,
there's no implementations of it.   We still need to do QA on our markup
though, which is why  I find the validator tool providing optional checking
of such things a valuable service (modulo the previously discussed bugs of
course).

> The HTML 4.01 specification imposes no requirements on error handling.
> If you ask me, that's a mistake, but that's the HTML tradition.

I have the complete reverse opinion, if the document is not valid HTML 4.01,
then it isn't an HTML document by definition, so the HTML 4.01 can't have
any say over what happens, and UAs should be free to do whatever they want
to attempt to fix it up and render something.

Jim.

Received on Thursday, 4 September 2003 22:11:28 UTC