Comments on: Any viable service registry should have UDDI as its foundation.

By: james governor

james governor — Tue, 16 Aug 2005 19:51:59 +0000

some notes on registry evolution from capience: http://vstory.techmarket.com/1/08/3/25423_3209272/article_7.html capience web site is awful though from a contact perspective

By: Radovan Janecek

Radovan Janecek — Tue, 19 Jul 2005 01:14:07 +0000

Hi James, good post. Please, take a look at my answer.

By: David Webber

David Webber — Sat, 09 Jul 2005 00:35:26 +0000

James,

I stated in January on the ebXML Vision dicussion list that Knowledge Driven Collaboration will begin to supplant WS-* as the way companies are planning, moving and adopting – and the OASIS ebSOA and BCM TC work is definately providing the basis for implementing that. Duane of course chaired the ebSOA TC and is now chairing the SOA RM TC – so no surprise that he also “gets this”.

By: miko matsumura

miko matsumura — Wed, 06 Jul 2005 10:30:55 +0000

Hi Byron, yes Systinet is experienced with UDDI, but Infravio has a lot of experience with UDDI in the context of integrating UDDI registry with JAXR registry repository.

This integrated approach came directly from our customer requirements and registry repository integration is being validated by the “Blizzard” platform as well as the Sun Registry.

We completely agree that more is needed than just UDDI (for example, repository), but of course UDDI is needed to ensure interoperability as it is a key piece of the Basic Profile.

Miko Matsumura
Infravio

By: Dennis Howlett

Dennis Howlett — Mon, 04 Jul 2005 21:57:59 +0000

I have to say I’m not totally convinced with the tagging argument – I’ve argued it both ways on my own site – Sig has interesting things to say about this at Thingamy but I still believe you end up with a mash up between tagging and some form of hierarchical repository. I can’t see any other way that works (but I’m not totally techie) and in any event current ‘search’ just doesn’t seem strong enough to do the bizzo.

Can’t say I’m totally enamoured with UDDI either. It doesn’t seem to have taken off in the way people expected. And in any event, isn’t UDDI meant to be an enabling thing rather than a core thing. Isn’t UDDI just search by another name but predicated on some person’s idea of what an electronic yellow pages ought to look like? Or am I being my usual simplistic self?

By: James Governor

James Governor — Mon, 04 Jul 2005 20:19:03 +0000

yes dennis i had considered that. thanks for the idea. i gently pointed to it above:

“Some of my views here tack into tagsonomy thinking, standing on the shoulder of the giant Clay Shirky. Ontologies have a place, but they don’t solve all known IT problems.”

By: Dennis Howlett

Dennis Howlett — Sat, 02 Jul 2005 06:39:38 +0000

Has it occurred to you that all these semantic problems can be solved with tagging? And without the baggage associated with structures and hierarchies. Just thought I’d throw it in…

By: Farrukh Najmi

Farrukh Najmi — Sat, 02 Jul 2005 00:59:58 +0000

There has been some interesting discussion on this subject that James and I have been part of at uddi-dev list. The thread begins at:

http://lists.oasis-open.org/archives/uddi-dev/200506/msg00005.html

It had a funny climax (if I may say so myself) at:

http://lists.oasis-open.org/archives/uddi-dev/200506/msg00031.html

Thanks as always for sharing your interesting analysis James.

–
Regards,
Farrukh Najmi

By: Byron Sebastian

Byron Sebastian — Fri, 01 Jul 2005 23:05:38 +0000

I tend to agree with you James, and I might go further. UDDI seems to be woefully inadequate in meeting customer needs, and I don’t know of many who are using it in a significant way. Having said that, Systinet seems to have as much or more experience than anyone else out there in terms of how registries are being used in the real world.