Re: Last try (really comments on recent ERB decisions)

At 5:19 AM 11/15/96, Henry S. Thompson wrote:
>I take it silence indicates agreement, that it IS too late to change
>anything, but being an ornery type, I'll try one last time:

Well, you got what you asked (too much of it in some cases).

>Will someone on the ERB stand up and be counted by explaining
>why an unprincipled hack (enumerating 11 grandfathered empty tags
>without amending the definitions of well-formedness and validity) is
>preferable to a principled hack (allowing  <?XML empty names ... ?>
>and defining its impact on well-formedness and validity, as in version

They are gone. This is the best of the decisions recently announced.

>Will someone on the ERB stand up and be counted by explaining
>why the built-in entity names are not overridable?

Well, they are now gone altogether. This is worse than having them not be
overridable. We have now ensured that HTML entities will not work without
elaborate declarations. We could have made the default entities
overrridable in the DTD subset, but that option does not appear to have
been discussed by the ERB (according to the posting). Needless to say, I
think this will damage XML acceptance by users.

    <anticipatory-comment>I'm sorry Len, but HTML is not an issue because
of vendors but because of users. SGML users transform documents all the
time -- that's part of what SGML is for -- HTML users do not, so they have
a greater claim for compatibility in my book.</anticipatory-comment>

>Will someone on the ERB stand up and be counted by
>explaining why being a proper SGML subset is being prevented by the
>enumerated attribute value change?

   It's been undone. However, we have now got the wrong decision and for
craven political reasons. We have endorsed a known design flaw, that WG8
already intends to remove, solely to avoid disturbing people by the timing
of our announcement. Since the SGML revision has been underway for some
years now, with few public announcements of work-in-progress, I am inclined
to think we may be justified in proceeding. Those of use who do not work
for DOD-Contractors organizations can't afford to wait forever.

   So, if we follow this recent revision, enumerated attribute values will
continue to be highly broken. I thought originally that the strict
SGML-subset idea should have been strangled in the grave, and now we see
the results in paying the price of pointless compatibility with SGML design
flaws that are already slated for demolition.

   -- David

I am not a number. I am an undefined character.
David Durand              dgd@cs.bu.edu  \  david@dynamicDiagrams.com
Boston University Computer Science        \  Sr. Analyst
http://www.cs.bu.edu/students/grads/dgd/   \  Dynamic Diagrams
--------------------------------------------\  http://dynamicDiagrams.com/
MAPA: mapping for the WWW                    \__________________________