[
Lists Home |
Date Index |
Thread Index
]
On May 6, 2005, at 5:58 PM, Richard Tobin wrote:
>> The point here is that a *syntax* defines nothing more than the
>> grammar
>> of a data format...
>
> And XML is not *just* a syntax in that sense. Some aspects of its
> semantics are made explicit in the XML spec, some are omitted, and
> some are left to the implementation.
Right... and this whole discussion started because of the lack of
clarity there. My assertion is that we'd be better off taking all
processing-specific notions out of the grammar, and specifying them
elsewhere... because ultimately, I can choose to ignore or implement
behaviour as I wish.
> People for whom whitespace between attributes is significant *in
> the application* are not using an XML application.
I think that's open to debate... I understand perfectly well where
you're coming from, it's just that in reality, few people make that
distinction. In a database that includes XQuery support etc. but which
also support the notion of "checkin/checkout", people *will* get upset
if things get munged much. It doesn't matter how many times you tell
them that this is correct behaviour.
|