Lists Home |
Date Index |
On Friday 18 February 2005 22:47, Elliotte Harold wrote:
> Prakash Yamuna wrote:
> > This becomes very useful from an evolution
> > perspective. The reason it is underspecified is a lot
> > of models have disparate needs and there has been no
> > common agreement on how they can expose it.
> This is backwards. The question is not what the models
> need. the question is what the engines consuming the
> source need from the models. If this can be standardized,
> then the models can provide it. For XPath and XSLT what
> the engines need is pretty well defined. For other
> uses--XPath 2, schema validators, XQuery engines--perhaps
> it's not so clear, but maybe we could come up with
The core issue for me here is that the processing software
needs to have a way of determining the capability of the
data model it is being asked to use so that it can adapt
its evaluation strategy according to those capabilities.
Without this we are forever stuck with having to use known
matched pairs of processing software and data model.
The options on the table are both trying to work
around this need either by requiring prior knowledge of the
specific format or forcing access via a known (and limited)
set of capabilities.
Using an adaptive strategy would be a better fit but I can't
quite convince myself that the extra complexity involved in
doing so will ever allow that to see the light of day, at
least in a general purpose API anyway.