Re: Spyglass HTML Validator

Peter Flynn <pflynn@curia.ucc.ie> wrote:
>[...]
>This is the problem with using libwww: it has never claimed to be an
>SGML parser, only to pick out tag-like objects from a data-stream
>purporting to be modeled on HTML.

	Yes, but my point is that during this "early phase" of
the Web's commericalization, which we all hope evolves into
a GoodThing, we should draw a distinction between the blatant
disregard for standards by a presently highly successful, initial
capital venture, and the evolution of HTML itself toward full
SGML conformance.

	The problems with "comments" are another example of this.
Originally, HTML had no concept of an SGML comment -- whatsoever!
A <!-- ...> was not a comment, but an "unknown tag" situation,
such that whatever lay between the '<' preceding the unknown
"element" and the next '>' was discarded (no checks for "--"
strings).  That's what "historical" comment parsing in RFC 1866
really means.  It's a euphamism for the original implementations
of HTML, by all WWW browsers.

	The first time anything resembling an SGML comment appeared
in HTML was with the SSIs for the NCSA server (by YouKnowWho 8-).
And <!--#foo ...> was not a comment for the then deployed browsers,
but, again, an unknown tag.

	Perhaps that's all just ancient history, but the Evolution
might be smoother and more harmonious if the target of full SGML
compliance were persued with a bit more appreciation that a lot
of what exists was done that way with good intentions by TheCreator,
and does not simply reflect the arrogance of YouKnowWho. :) :) :)

				Fote

=========================================================================
 Foteos Macrides            Worcester Foundation for Biomedical Research
 MACRIDES@SCI.WFBR.EDU         222 Maple Avenue, Shrewsbury, MA 01545
=========================================================================

Received on Monday, 28 October 1996 18:40:22 UTC