Jul 13, 2009
Participants: Barseghian, Bowers, Crawl, Gallagher, Hosseini, Ludaescher, Riddle, Wang
- parviez - marratech 'chipmunks' problem, restarting doesn't help. barseghian - reported (this is 3rd time, peter and james both had this issue before) to thomas and ron.
- barseghian - working on highlighting failed runs in WRM, finish delete from provenance code, working on EI paper revision.
- ludaescher - what signifies a failure? Many different definitions. barseghian - get everyone to start using Throw Model Error?
- supriya off for the summer, but hope work on aggregation will continue in future
- engineering view? barseghian - not yet started, crawl finishing off some provenance issues
- riddle - working on tagging, could use some testers now. Tags being saved in moml.
- tags discussion
- any named object can be tagged.
- wrm will be able to register with tagging so users can tag runs. Put tags column back in the future so we can search them.
- TODO shawn to coordinate with aaron re: annotation and saving into library problems
- TODO tagging team to check with christopher re: a system-wide selection listener for tagging everything
- wang - presented distributed usecase reap paper at ICWS in LA. Some other groups also using Kepler.
- ideas for extending distributed execution beyond embarassingly parallel computation.
- gallagher - getting ready for a new release of the server, a plug-in module that existing users can just drop in to use. This will enable some of the EML stuff discussed previously.
- discussion of search interfaces. Where should search for e.g. OPeNDAP and DataTurbine go? No registries to search. Data tab vs in the actor vs specialized data search actor.
- TODO look over terrestrial usecases, update with Status
- bowers - will we present our work to usecase creators? barseghian - original plan was for a final REAP meeting where we do this with the original group, TODO check with matt
- Bertram will arrange next week's meeting.
|