Re: Dropping the Normative Reference to SGML (was: I-D ACTION..)

On Wed, 6 Oct 1999, Rick Jelliffe wrote:

> ISO 8879:1986 as corrected by Annex J,K,L does allow restrictions to be
> placed on the kinds of SGML that are permitted in a particular document
> type.  This uses the SEEALSO parameter of the SGML declaration, and is
> explained in the section on "Added Requirements".  See
> http://www.ornl.gov/sgml/wg4/document/1960.htm

Well, my argument on this particular point is based on 15.2.1 and 15.2.2
of ISO 8879.  The "shall not prohibit any markup that this International
Standard would allow in such documents" is of course subject to the
specifics of an SGML declaration for what in fact is "allowed" to begin
with, and we now have a lot more wiggle room.  Now, K.3.9 says

: The public identifiers identify additional requirements for the
: document, including requirements unrelated to the SGML language. These
: requirements are in addition to, and must not contradict, the
: requirements of this International Standard. 

which suggests no new interpretation of 15.2.2.  And none of the examples
in Note 22 suggest that markup constructs are also in play (which is
basically what 15.2.2 is about.)

But, yes, Annex L does mention restrictions on "SGML language constructs"
for XML. (Stuff like data attributes and &-models, I suppose)

> [...], it is perfectly legitimate to make a WebSGML application which
> does not allow internal subsets, as an added requirement.  

Unfortunately, both reasons for wanting this - the overt one ("we need a
version indicator, so we nominate the DTD FPI for this") and the tacit
one ("Keeping The Web Safe For Netploder, since it'll barf on anything
else") - are bogus.  Versioning is still a *real* problem, and the plight
of Netploder is no excuse for establishment via a formalistic fig-leaf.

> (I proposed and drafted the SEEALSO bits, but this may indeed be
> something that could be clarified in some future rewrite of SGML if
> anyone wants that.)

Perhaps a clarification with respect to 15.2.2?  Other than that, how
about a detailed write-up of the SEEALSO parameter (i.e. motivations,
scenarios, substantive intent, etc.)?  Actually something of this sort for
all the new WebSGML stuff would be a welcome addition to Wayne Wohler's
classic tutorial.

> I agree with Arjun that HTML may not be "Conforming SGML Application",
> but it certainly is an "SGML Application"  with specific Added
> Requirements.

As a design document (e.g. DTD(s) plus documentation for an enabling
architecture), yes.  As a "standard" aimed at real-world practice, defined
as stupefied awe of hysterically hyped wowsers, no.

Another way to draw a distinction is to say that East is "HTML, the SGML
application" and West is "text/html, the bits on the wire", and never the
twain shall meet.  An SGML application with a raft of scraped and
scrounged "Added Requirements" is not the way to describe, much less
rationalize Web practice. 

> So I think it would be slack to remove normative references to SGML:
> they can be updated to WebSGML, since that addresses several 
> requirements from the HTML world 

Like it or not, KTWSFN is an integral part of W3C activity.  I don't think
I have to remind you of smug officious sanitized euphemisms like "Level 3
browsers";) 

It's time to call a spade a spade.


Arjun

Received on Wednesday, 6 October 1999 12:35:42 UTC