W3C home > Mailing lists > Public > www-tag@w3.org > June 2004

RE: xml11Names-46

From: Jacek Kopecky <jacek.kopecky@deri.at>
Date: Wed, 16 Jun 2004 09:51:41 +0200
To: Joshua Allen <joshuaa@microsoft.com>
Cc: Elliotte Rusty Harold <elharo@metalab.unc.edu>, www-tag@w3.org
Message-Id: <1087293924.2721.12.camel@Kalb>

Hello all, 

first of all, because the problems of moving from XML 1.0 to XML 1.1 are
technical and the decision to issue erratas or to create new versions of
the specs spans the W3C and again, is technical, not a question of
process, I do believe the TAG has a say here.

In reply to all the cries of horror 8-) about attempting to rewrite
history:

I acknowledge there are incompatibilities between XML 1.0 and 1.1,
that's why the version number has changed. I believe we should plan for
the long term here, though:

In the future, I expect only one of the two version (unless a next
version comes out) to be widespread, optimally that should be XML 1.1,
but it can well be that (almost) nobody actually goes and uses XML 1.1
so it would stay to be XML 1.0.

In the future, I expect most parsers to be able to handle both versions
of XML, with no substantial changes to the APIs. For example DOM in Java
will not change, will it? Will DOM or SAX change in any language,
actually? Will the Perl API change? These expectations I make based on
my optimistic view of the affairs.

Some applications may require XML 1.1. Why should any application
require XML 1.0 and *forbid* XML 1.1, though?

In the long term, I believe that an errata to a spec (say XML Schema
1.0) is better than having two XML Schema specs. There will be a log of
changes, I don't suggest we rewrite the Rec, I suggest we issue a proper
second edition. Is that a re-write of the history? I don't think so.

I just think the benefits outweigh the problems by a huge lot. My
opinion, mind you. 8-)

                   Jacek Kopecky

                   Digital Enterprise Research Institute
                   http://www.deri.at/




On Sun, 2004-06-13 at 18:37, Joshua Allen wrote:
> > Here, APIs for XML already use UNICODE strings to represent the Names
> 
> What?  You have a list of important implementations and their
> implementation details?
> 
> > an XML parser and serializer, I would just use UTF-16 or whatever to
> 
> Oh, I see, you are speculating based on what you think you might
> possibly do if you wrote your own parser.
> 
> > I don't even think DOM should be changed, for instance, except in
> parser
> > and serializer parts where the DOM implementation should be able to
> > indicate supported versions, if that is not already present there.
> 
> Great; we have literally millions of developers depending on shipped
> code which implements these specs.  Let's just *assume* that all of the
> implementations do things a particular way.
> 
> > Rewriting history is bad when you change the intents, but I don't
> > believe it was the intent of XML Schema 1.0 to limit itself to XML
> 1.0.
> 
> I hope to God it wasn't their intent to have the scope apply *beyond*
> XML 1.0, since there were no other versions of XML at the time.  And if
> that was their intent, they failed anyway.
> 
> I think it's a profoundly bad idea to rewrite history by pretending that
> specs which have shipped to millions of people are really meant to
> support a new spec which breaks backwards compatibility.
> 
> 
Received on Wednesday, 16 June 2004 03:51:58 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:26 GMT