Re: well-formedness error

[Mark Pilgrim <pilgrim@xxxxxxxxx>]

On Tue, 22 Jun 2004 12:43:11 +0900, Martin Duerst <duerst@xxxxxx> wrote:
> 
> At 12:58 04/06/19 -0400, Mark Pilgrim wrote:
> 
> >On Sat, 19 Jun 2004 18:10:42 +0200, Julian Reschke
> ><julian.reschke@xxxxxx> wrote:
> > > >>A perfect XML parser is one that accepts wellformed documents and
> > > >>rejects all others. So from that point of view, for instance MSXML is
> > > >>perfect (or close to that).
> > > > http://www.intertwingly.net/blog/1689.html#c1074118904
> > > I talked about MSXML. You are referring to a comment on .NET's XML
> > > parser. These are different things.
> >
> >Sorry, I meant http://www.intertwingly.net/blog/1689.html#c1074112699
> 
> Is this a real example, or a constructed one?

It is a real example from a real feed.  The character was unknowingly
copied-and-pasted into a blog entry and ended up unaltered in the XML
feed.  Sam noticed it and made a mirror of it.

> And if an 'unknown glyph'
> box is displayed, that may well be enough for motivating the author
> to fix it.

That's not the point.  The point was that the version of MSXML which
ships with IE 6 is not a conforming XML parser, since it does not
reject that non-well-formed feed.

As Julian has pointed out, later versions of MSXML fix this bug. 
Perhaps IE 7 will ship with a conforming XML parser.

As I have pointed out, even the latest version of MSXML is not a
"perfect" XML parser (defined by Julian to mean "parses all
well-formed feeds and rejects all non-well-formed feeds"), since it
does not respect RFC 3023.

-- 
Cheers,
-Mark