Conversation and restart

[Norman Gray]

Folks,

We had a skypecon this morning. This was necessary to restart us, as
we seem to have lost a bit of momentum since the beginning of the year.

Overall, we're in a not-bad state. Qsac has all its major
functionality, though there are still gaps which we should plug with
specific user-stories; Spacebook has its major functionality, though
there are features Tony still wants to add; the client library does
what it says on the tin, and is in a state where it could be adapted
and extended as appropriate.

So: what next.

* We're now in a position where we can usefully start tracking bugs
and user-stories. Norman will adjust the SKUA googlecode issue
tracker to help this.

* Kona will return to looking at VOExplorer, with a view to extending
the current preliminary SKUA support into something useful. With that
in mind...

* Kona wants a user-story for VOExplorer searching

* Tony raised performance issues: (i) is the current Jena-based
triplestore powerful enough?; (ii) we anticipate that there will be a
performance hit from repeated delegation of queries, and we should
test performance with physically separate SACs and add the
functionality to let users make no-delegation queries. Some
discussion: true, but Norman is nervous about adding features or
tuning too early, and would prefer to wait until we have a measurable
performance problem in each case. Changing Quaestor to use an
alternative triplestore would be feasible, but not trivial; also, it
seems that the Jena folk at least _intend_ TDB to be competitively
performant.

* Version of TDB. Norman should check that Quaestor still works with
the most recent versions of TDB.

* apps.myskua.org: this is living within an existing Tomcat, which
Kona needs to bounce occasionally; also, this can't be migrated to
Java6, which TDB needs. Thus we should either have a separate Tomcat,
or else we can use the standalong Qsac. Problems with the latter: (i)
there is some high-level documentation and a webpage which is served
by the current Tomcat separately from the Qsac application; (ii)
having a separate server might involve some routing configuration.
Kona will investigate with MarkH -- ideally, http://apps.myskua.org:80
would resolve to the standalone server, if necessary via some router
magic.

* We'll aim for another iteration restart on 23 March. This is after
the Semantic Astronomy Workshop, and after the AIDA Strasbourg meeting.

* Tony mentioned that we should check with JISC when it is they want
the SKUA final report.

That's us, I think!

All the best,

Norman

--
Norman Gray : http://nxg.me.uk
Dept Physics and Astronomy, University of Leicester