Who gets the credit (or blame) for Web Service / Service Oriented Architectures

The SOA blog asks the interesting question Who gets the credit (or blame) for Web Service / Service Oriented Architectures with reference to Dion Hinchcliffe's post on the subject. Probably this is a question that can't be fully answered since at least in the case of Software Oriented Architectures the overall vision is pretty generic and the influences many (perhaps for Web Services per-se its easier - although arguably even there a great deal is borrowed from precursors such as OMG's CORBA). Here's my post on Dion Hinchcliffe's blog:

The timeline provided here is a good chart of technologies - however it seems to me that you have to look a lot further back for the core concepts/paradigms of Service oriented Architectures. Work in Agent technology research (e.g. see conferences such as AAAI, IJCAI and AAMAS amongst others) has long maintained a view of computing systems consisting of collections of distributed service provider/consumer systems. This is likely true of other fields such as distributed systems.

From 2000 onwards for example the Agentcities european projects (http://www.agentcities.org/EURTD) build up a large scale testbed o automated service providers / consumers running in a public/open environment which were able to discover one another, interoperate and form applications. The largest demonstration took place in 2003 (http://www.agentcities.org/note/00001/actf-note-00001a.pdf) but the first messages in the network were sent in 1999/2000. The technologies were not WSDL, SOAP etc. (they used specifications from the now little know Foundation for Intelligent Physical Agents - FIPA / http://www.fipa.org) - but the concepts essentially the same as current SOA models.

The current W3C Web Services Architecture bears a lot of similarities to the FIPA Abstract Agent Architecture (published in http://www.fipa.org/specs/fipa00001/ 2000-2001) - not entirely surprising since there are some common authors.

And probably we need to go back even further at higher levels - such as to Hewitt's Open Systems papers.