OASIS Search Web Services TC by vev19514


									                          OASIS Search Web Services TC
                                 October 3, 2007
                                 Meeting Notes

All are voting members.. 5 of 7 voting members present, quorum.

Discussion of the strawman

- Boilerplate. Discussion of whether to leave it in or not, just for the strawman. Leave it in.

- Overview. Jan will write it.

- Delete section 2. Merge "Search Web Service Overview" into section 1 and renumber.

- Data model. There is a stub for "Introduction, Overview, and Data Model" at 5.1. Section 5 is
Transport. So the question is, is there a data model in mind for transport. Matthew had put that
in, and he's not on the call. A more general question is, should there be a data/information model
at the beginning of the document. Probably so. So strike 5.1 and add it at the beginning.

- Operations. Kerry questions whether we are using "operation" correctly. Need to discuss this
via an email thread. We think we're using it properly in the web services context, but this use
came about in the SRW context, and now that the protocol is SRU oriented maybe we need to
revisit. Kerry feels that the term "behavior" is more common, at least in an SOA context.

-Shiboleth/SAML. Kerry will try to supply prose about possible use of Shiboleth/SAML to add to
the (non-normative) annex on authentication.

- Opensearch. Will be added. See below.

- Next steps. Revision in two week, meet again in three. Will completely rebuild the document as
editing it has become near impossible because of the formatting. Will maintain a master version
(possibly in wordperfect) and convert each time to pdf (and possibly word). Meeting in three
weeks, we will hopefully approve the document as a strawman (not a committee draft) and
announce it for public review.

The strawman document will include an annex that describes how the protocol is to be used for
openSearch requests and responses. TC feels it is important that the initial public submission
address openSearch concretely, that that will get peoples attention. And this will be a "strawman",
which means it can be "shot down". Even if we don't get it right in the strawman, it will alert
people that we are serious about it.
The annex will actually replicate the openSearch spec and discuss the relationship to the
protocol. It will make it clear that you can craft an openSearch request that is compliant with the
SR protocol, and that any SR server can be openSearch friendly. However it will not concretely
address the question of whether any openSearch request is necessarily SR conformant.

responseSchema Parameter
This necessitates a modification to the protocol: addition of a parameter that says what schema
the response comes back in. Possible values: (1) SR2.0 (default) (2) atom1.0 (3) rss2.0 (4) html
(?) . Adding this parameter does not cause incompatibility because it can be omitted (defaults to

Publicity, Recruitment, Etc.
When the strawman document is announced, that will be the time to further pursue recruitment.
That may also be a good time for a press release. OASIS likes to have three sponsor members
before putting out a press release. Booz Allen Hamilton and IBM have recently joined and both
are sponsor members, and along with DEST, we now have three. We'll discuss with OASIS
putting out a press release in conjunction with the release of the strawman.

OpenURL, Holdings, Etc.
We deferred discussion of this for the most part, noting that there are three potential context sets
for consideration that are realted in one way or another to holdings:
1. A basic holdings set, which might consist only of identifier, location, and availability indexes.
2. Detailed holdings (serials), suggested by Ralph.
3. Enumeration and chronology. Jan suggests enumeration and chronology is more at the
resource description level than at the holdings level.

Use Cases
We deferred discussion of use cases. Sophie had suggested a use case for mashup
components, which sounds interesting, but she was not on the call.

Next Meeting
October 24, 9 eastern US. Hosted (again) by Ralph.

To top