RE: URL better than FPI

> From:	Murray Altheim [SMTP:altheim@eng.sun.com]
> 
> I don't follow the logic at all. XML can reference ISO 8879 without
> including the text of it, nor is it enlargened by such a reference,
> 
	I said "include by reference", which doesn't mean
	including all the text, however it does imply 
	including all the complexity.  One of the main reasons
	for having XML, rather than simply using SGML, was
	to have a standard that more people could understand
	and work with.  Having the standard defined by a 
	small document, plus references to character tables
	is intended to achieve this.

> and whether it costs money to purchase a copy of a spec has 
> absolutely no relevance on whether or not it is normative to XML. 
> 
	It's an unfortunate fact of life that most things written
	to standards are written without sight of the standards.
	Real life HTML is an example of the result of this.  Making
	standards available without cost significantly increases the
	number of people who actually write from the standard, rather
	than folklore (although HTML demonstrates that the majority 
	may still rely on folklore).

	I'd suggest that a lot more people have read the Email and
	MIME RFCs than have read the X.400 reccommendations.

> I believe one of the reasons why ISO 8879 is not in the normative
> section (but is included in 'Other References') is perhaps because
> at the time of printing the WeBSGML (TC2) was not yet an ISO standard
> 
	I thought we were talking about SGML - XML claims to
	be SGML, although any transitive reference to SGML
	would still introduce the full complexity.

> and therefore couldn't be referenced as a normative specification. 
> But I don't remember the particular history on this decision.
> 
>

Received on Monday, 21 February 2000 14:52:14 UTC