[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: "Rick Jelliffe" <ricko@a...>
• Date: Sat, 27 Jul 2002 00:55:24 +1000
• Cc: <www-xml-blueberry-comments@w...>
• References: <200207261437.KAA25077@m...>

From: "John Cowan" <jcowan@r...>
 

> Rick Jelliffe scripsit:
> 
> > Why not try that minimal solution first, and see if it works? 
> > 
> > (And why not do it as a corrigendum rather than 1.1)
> 
> I keep telling you why.  Because it changes the definition of well-formedness.
> A document like "<foo<NEL>bar=baz> is not WF XML 1.0, but it is WF XML 1.1.
 
And it does not matter.  If it made any well-formed document not-well-formed,
it would matter.    Currently there is no guarantee that if you send a document
in EBCDIC that it will be accepted by an XML parser.  All this means is
that, for the transition duration, if you use EBCDIC NEL conventions you
still have no guarantee: you will need to send to friendly systems or generate
friendly newlines.  Anyone who needs to send NELs in data but not have
them mapped to #A can continue to do so: they just need to use character
references, so there is no change to the kinds of data that can be sent over
the wire. 

This absense of NEL is an error in XML 1.0 that should be corrected in XML 10. 
That it expands the definition of well-formedness is a good thing, not an argument
against it. 

Cheers
Rick Jelliffe

• References:
  • Re: version numbers and infosets
    • From: John Cowan <jcowan@r...>

• Prev by Date: Re: URIs harmful
• Next by Date: RE: sax parser and xpath
• Previous by thread: Re: version numbers and infosets
• Next by thread: RE: version numbers and infosets
• Index(es):
  • Date
  • Thread