[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: version numbers and infosets

• To: xml-dev@l...
• Subject: Re: version numbers and infosets
• From: Richard Tobin <richard@c...>
• Date: Tue, 23 Jul 2002 10:43:24 +0100 (BST)
• Cc:
• In-reply-to: <003c01c231eb$96f290f0$fe193044@tbp1>
• Organization: HCRC, University of Edinburgh

In article <003c01c231eb$96f290f0$fe193044@tbp1> Tom Passin wrote:

>Richard, would you clarify this point?  I saw that the production required
>"1.0", but withou the requirement to produce an error, my own interpretation
>would be that a processor could accept, say, "1.1" and proceed without an
>error.

I think that is the best solution - see below.  But:

> Was the idea that the production would be so definitive that
>obviously a well-formedness error should be produced?

Yes.  But I have since been persuaded that it would be a mistake to
require 1.0 processors to reject 1.1-labelled documents that are
otherwise well-formed XML 1.0.

>The rationale for the change was said to be XML 1.1, and personally I think
>it should not have been published until 1.1 was published - they should have
>been coordinated.  What if 1.1 never happens in the anticipated form?  What
>of the change to 1.0 then?

Yes, it was intended to pave the way for XML 1.1.

If there was never a new version of XML, it wouldn't really matter
what 1.0 said about version numbers, because there would never be any
later-than-1.0 documents.  If there was a new version that was quite
different from the proposed 1.1, then that might have been a problem.

I must say that some of the reactions to this seem excessive.  Does
anyone have any real documents with a version number later than 1.0?
I don't think so.  No real documents are broken by any change
concerning versions other than 1.0.  This is always the most important
point when we are considering errata.  Whether implementations have to
change is important, but less so, because there are far fewer
implementors than users.

I think the best compromise is to say that documents labelled with a
version other than 1.0 are not well-formed XML 1.0 documents, but that
an XML 1.0 parser may accept them.  This breaks no documents and no
implementations, but it allows us to unambiguously state the version
of a document.  Can anyone see any problems with it?

-- Richard

• Follow-Ups:
  • Re: version numbers and infosets
    • From: "Thomas B. Passin" <tpassin@c...>
  • Re: version numbers and infosets
    • From: "Karl Waclawek" <karl@w...>
  • RE: version numbers and infosets
    • From: "Michael Kay" <michael.h.kay@n...>

• References:
  • Re: version numbers and infosets
    • From: "Thomas B. Passin" <tpassin@c...>

• Prev by Date: RE: [ANN] Ipedo XML Database 3.1 Now Available
• Next by Date: RE: version numbers and infosets
• Previous by thread: Re: version numbers and infosets
• Next by thread: RE: version numbers and infosets
• Index(es):
  • Date
  • Thread