Re: Interpretation of RE

Craig Hubley (craig@passport.ca)
Wed, 12 Apr 95 04:49:47 EDT

> At 04:46 PM 4/10/95 -0700, Arthur van Hoff wrote:
>
> >> This must be based on clause 7.6.1 of the SGML standard, which states "If an
> >> RS in content is not interpreted as markup, it is ignored."
> >
> >That is where I got my information from. This rule is ignored by most
> >browsers inside a PRE element though. Should it be?
>
> Not as far as I'm concerned. Most HTML product vendors don't take the SGML
> standard seriously, probably because most web info providers are not much
> concerned with document interchange, structure driven search, etc.

Sure, but every site that has a significant document management requirement
is taking is *very* seriously because SGML is the 'maintenance format' for
these documents. The long term cost of maintaining data directly in HTML
is horrendous and I predict that it will break any organization that tries
it. The SGML-automated alternatives are far less expensive to run long term.

Anyway, I'd like to assemble a list of HTML products that disobey SGML rules,
possibly to arrange a boycott/smear-campaign/warning-label program of some
sort. Anyone that has any such information please forward it to me, or if
there is some repository of this information already, please let me know.

This could help the committee determine just how badly violated a particular
SGML tenet is, perhaps determine why, and propose an SGML-compliant solution
to the underlying problem/requirement. Then there will be no excuse for the
laziness and we can keep this flailing horror called the 'Web' on track to
become something useful for more than just printing badvertising.

> >What is nsgmls?
>
> An application of James Clark's sp parser, and a successor to "sgmls". It's
> freely available at places like the SGML archives at ftp://ftp.ifi.uio.no.
> I use nsgmls to validate HTML as SGML and translate the parser output into
> "browser HTML".

The main reason to stick closely to SGML is to leverage the existing parsers,
editors, industry-specific DTD standards, etc... if HTML deviates from this
it will be ten years before a credible hypertext knowledge maintenance
standard evolves. It is tough enough now just trying to figure out how to
maintain names of documents, with HTML anchors containing URLs working now,
more robust anchors containing URI/URNs on the way, and HyTime links possibly
coming down the pipe sometime.

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