[
Lists Home |
Date Index |
Thread Index
]
At 03:55 PM 7/6/2002 -0400, Jonathan Robie wrote:
>>That's easy. Layer XML Query on the Infoset, drop the PSVI, and you have
>>half a chance of producing something likely to have a long and useful lifetime.
>
>That's not layering, that's deciding that data is not typed unless I
>require explicit casts, convert it to Java or relational data, or
>whatever. And that makes many simple examples come out wrong, eg:
>
> let $x := <foo xsi:type="decimal">39.42</foo>
> let $y := <foo xsi:type="decimal>147.23</foo>
> return
> if ($x < $y)
> then "x is less"
> else "y is less"
>
>Your approach would mean that the above query would return "y is less".
>Trust me, many users of XQuery would not like that.
They can have their PSVI Query, then, and stay away from XML Query.
Alternately, they could ingeniously rewrite that as:
if (num($x) < num($y))
and get the correct result. I see no advantage to your proposal.
>>Otherwise, change the name to PSVI Query and let us XML folks get on to XML.
>
>XQuery and XPath have created a simple, typed data model for XML - I think
>this is a Really Good Thing [1].
I think this is a very bad thing, but I'm growing very tired of arguing
with people who want and want and want more features.
>XQuery is defined in terms of this data model, not in terms of the PSVI,
>which was created for different purposes and is really on the wrong level
>of abstraction. There is a mapping to this data model from the PSVI, and
>mappings in the works for at least DTDs, well-formed XML, and XML views of
>relational data.
Sadly, that model is built on notions which are much like those of the
PSVI, not a simple XML 1.0 Infoset. That it isn't built directly on the
cracked foundations of W3C XML Schema isn't much consolation.
Simon St.Laurent
"Every day in every way I'm getting better and better." - Emile Coue
|
