RE: Bad News on IE6 XML Support

[Benjamin Franz <snowhare@nihongo.org>]

On Sun, 9 Sep 2001, Joshua Allen wrote:

> >or Sun or IBM or Apache or one of the free tools.  An editor is a tool
> >for correcting mistakes; a browser does not provide any mechanism
> >for correction, so the correct action of a browser (and a parser, in
> its default mode) is to reject the file.
>
> This is getting pretty ridiculous.  IE is not an XML parser, period.

[snip]

I'm starting to wonder if this isn't just another example of MS's infamous
'Embrace and Extinguish' policy in operation. XML works well (by *design*)
on any OS and any computer language.  It is open and available to all
comers. It is being extremely successful in creating services that are not
tied to any one platform. 'MSIE-XML' only works on Microsoft's software
platforms but is trumpeted (at least in all the places the consumer is
likely to see) as XML.

The _specific_ reason the XML specification says 'scream and die' on parse
failures is *because of* Netscape's and Microsoft's web browsers being
'liberal' in 'parsing' HTML. One of the explict _goals_ of 'scream and
die' was to prevent that from happening, again, in the browsers. For
Microsoft to now say 'But MSIE isn't an XML parser (it just quacks like
one most of the time)' as justification for breaking that _specific_ goal
of XML is dis-ingenuous at best.

The _RIGHT_ thing, the thing that doesn't fragment the world into
'Microsoft applications' vs 'everyone else', the one that _minimizes_ the
damage from non-XML docs running around saying '<?xml version='1.0'?> is
to _STOP_ the clear problem being perpetuated while is is still
relatively small. Right now.

The _WRONG_ thing, the thing that divides the world into apps that work
only with "MSIE-XML" (an underspecified thing defined by "what Microsoft's
software accepts" and subject to change without notice), the one that
spreads unlabelled 'MSIE-XML' so wide and far that the previous assumption
that things labled 'XML 1.0' could be parsed by _anyone_ compliant with
the XML 1.0 spec is utterly obliterated, the one I will be disappointed
(but, unfortunately, not suprised) to see is the continued 'adjustment' of
Microsoft's software to generate and read "XML" that _ONLY_ Microsoft's
software can parse. For the sake of 'compatibility' _with software
Microsoft itself caused to be non-compliant_.

Any bets as to whether the generic MS-XML parsers will have _their_
defaults altered in the future for the sake of 'compatibility' with the
_intentionally_ broken behavior of IE wrt XML?

-- 
Benjamin Franz

To paraphrase Russ Allbery,
   "..by God I KNOW what this (XML) is for, and you can't have it."