RE: [xml-dev] processing instruction with 'xml' target

[<juanrgonzaleza@canonicalscience.com>]

I have replied several "you are right", "this is right" to contradictory
points. Still nobody noticed it doing this thread less funny it would have
been when noticed. In the end, the X in XML is for eXciting!

I am not sure if when I write <?xml version="1.0"?> in a doc I am writing
a xml declaration or a PI. Yes, I have carefully read Lee, and David, and
Tim, and many others and also the spec (in several versions) and respect
their opinions but still I am not sure because of next argument.

The reply to above question depends of the arbiter used, the formal spec
or a real working implementation?

Apparently my Mozilla based browser sometimes deals the xml declaration as
a PI. I obtain something like

<message>
xml processing instruction not at start of external entity
</message>

Also this source from Microsoft states that XML is a PI for the Internet
Explorer –at least in the CDF-:

[http://msdn.microsoft.com/workshop/delivery/cdf/reference/xml.asp]

I have read of some SAX implementations doing the same.

Therefore far from Mike’s absolute verdict <quote>They are factual
statements which are either true or false, and can be verified by
reference to the specification</quote> I agree on that <?xml
version="1.0"?> is not a PI in the spec but _could_ be a PI in some real
implementations FAPP. Sorry to say this but I never work with the formal
spec, just with real implementations in browsers, tools...

<quote>But in most cases it really doesn't make any difference whether it
is or not, so feel free to look at it as one if you wish.</quote>

I like the pragmatic attitude of Kurt Cagle rather than the chic formality
of Mike here.

P.S: A theoretical discussion about if a real life implementation is not
100% compatible with the formal spec would be or not be called an
implementation of the formal spec is of no real interest here. I work with
real implementations either if you call it "xml implementation",
"pseudo-xml", "partial implementation of the 1.0 spec" or any other. It is
just a naming question.

Juan R.

Center for CANONICAL |SCIENCE)