On Fri, 7 Oct 2005, Howard Noble wrote:
> c.. I see that IESR is doing something very similar to the pilot we
> are working on: http://iesr.ac.uk/use/ and wants to use UDDI (but
> Matthew Dovey is working with these groups anyway)
"Wants to use UDDI" is probably a bit strong. The IESR is currently
evaluating UDDI (and a report should be available in due course).
Depending on the outcome of that report, it may implement UDDI, it may
not. At least, that is my understanding.
From a very personal perspective, I have doubts about whether UDDI is the
right way to go, though I recognise that it does seem to be the solution
of choice within the UK eScience community??
What is clear, is that we want to build up a distributed network of
cooperating service registries - global in nature and crossing various
domains. In some ways, service registries may become a bit like the DNS -
though that analogy has some weaknesses so I won't push it. But it seems
to me to be a good thing if we can push ownership and maintenance of
knowledge about services right out to the service providers, in the way
that DNS tends to do.
On the face of it, UDDI looks like the right choice for doing this. That
is what it was designed to do after all. But on closer inspection, we
find a relatively complex standard, that is linked quite closely to SOAP
(at least in practice, if not in theory) and that offers an overly rigid
model in terms of the kinds of things that can be described.
In short it isn't very flexible - as opposed to, say, OAI-PMH, which takes
the more open-ended approach of offering a framework into which you can
pour almost anything.
As a community, I think we will get further, faster which an approach
based around exchanging service registry records using the OAI-PMH - this
is essentially what the IESR offers now, and is also the approach adopted
by the OCKHAM project in the US.
As I say, these are my personal views - as someone not very tightly
involved in the IESR project. The study currently being undertaken by the
project will, hopefully, help to inform our discussions and decisions in
this area. And of course, these kinds of decisions are not solely about
the 'best' technology - but are dependant of what other parts of our
community and the communities around us choose to do.
Andy
--
Distributed Systems, UKOLN, University of Bath, Bath, BA2 7AY, UK
http://www.ukoln.ac.uk/ukoln/staff/a.powell
tel: +44 1225 383933 msn: [log in to unmask]
Resource Discovery Network http://www.rdn.ac.uk/
|