[
Lists Home |
Date Index |
Thread Index
]
John Cowan <jcowan@reutershealth.com> wrote:
[Eric van der Vlist:]
|> On Sat, 2002-09-21 at 01:07, Arjun Ray wrote:
|
|>> AFAICS, a parser should cover everything in the infoset.
|
| Note that what got into the infoset was a purely arbitrary choice
| by yours truly, with an eye to SAX, DOM (much diminished later) and
| XPath. There is nothing normative about the selections,
How mutable is the spec, then?
| and saying "but it's in the infoset!" is no sort of reason why a
| parser ought to support something. Infoset conformance depends on
| documenting what you in fact do support, neither more nor less.
Yes, but in what terms is the documentation expressed? Is there an
"infoset-speak" behind the scenes?
|> infortunate that there is no XML serialization of the XML infoset (and
|
| The serialization of the XML Infoset is XML.
But not necessarily the original document, right?
|
