[MOBY-l] UDDI vs. WS-Inspection vs roll-your-own

[Phillip Lord]

>>>>> "Mark" == mwilkinson  <mwilkinson at gene.pbi.nrc.ca> writes:

  Mark> Phillip Lord wrote:
  >> I guess my approach to experimentation, is use a standard until
  >> you know that it is broken, or just will not do what you want it
  >> to.

  Mark> Possibly.  It seems more likely, however, that you will limit
  Mark> your thinking to only those things that you know the standard
  Mark> can handle ;-)


Actually I think I would agree with this!

  Mark> In any case, I simply don't care.  Bah!!  What we're trying to
  Mark> accomplish here is the definition of a *behaviour* that allows
  Mark> us to discover biological data and services in an intuitive
  Mark> way, and hopefully facilitates the ability to build fully or
  Mark> semi-automated systems to do this for us. All we need to
  Mark> determine right now are the components of the system, and the
  Mark> set of API calls that will accomplish this.  After that, you
  Mark> can implement it any way you please - if it can be done on top
  Mark> of UDDI, then fantastic.  If not, then we either suggest to
  Mark> the UDDI consortium that they add new features, or we stick to
  Mark> our own registry system (or we use whatever new standard has
  Mark> appeared in the meantime...).  Either way, I simply don't care


I can also appreciate this. 

I think that this reflects our slightly different concerns. As you
say, Moby is mostly about a transport layer. While this is an issue
for mygrid, for me personally, I am more interested in the metadata
issues, and I just need a transport layer there to allow me to do this
sanely. The work that Robert did on Tambis, for instance, is a nice
example of this. Good, interesting and important work, but much harder
to do because of the lack of a good transport layer. 


  Mark> - I just want to figure out how to make it do what we all want
  Mark> it to do as quickly as possible, and we can experiment with
  Mark> various underlying implementation issues once we get there.
  Mark> The critical thing is to determine which of our ideas are
  Mark> well-founded, and which are an ugly can of worms - personally,
  Mark> I think I can do that faster with the very simple registry
  Mark> that we have running already... 


Again I can see the logic. I think having Moby, use MOBY central,
while mygrid is doing more UDDI related things, provides us with a
good opportunity to be honest to test out the differences. 

I would certainly agree with you that moving Moby toward a UDDI based
registry would not be a sensible use of time. You have Moby
Central. It is up, and it is running. 


  Mark> Phil, we need to go for another beer!

My liver hasn't recovered from ISMB yet, and my digestive
system is still in shock over the Canadian approach to vegetarianism
(you get a potato with your grilled cow). I'm sure that more beer will
happen soon though!

Cheers

Phil