|
Aug 17, 2009
Participants: Barseghian, Crawl, Jones, Ludaescher, Riddle, Schultz, Wang
- New oceanography usecase
- Crawl: workflows for edge detection on different SST datasets -- Peter has prexisting code (fortran, matlab, maybe python) for edge detection.
- Jones: will this contribute enough to REAP to be worthwhile in the remaining time?
- arguments for: will likely make the Matlab, External Execution and Python actors more robust. Good additional demos to other oceanographers
- will revisit once we have more details from Peter
- engineering view mockups
- Jones: flesh out usecases, add usecases, reorganize page so it's apparent what these proposed features will do to address the needs of usecases, once further along then present to Eric
- workflow run manager
- Jones: finish it enough so it's useful and stable for 2.0 release
- Derik: ok, hopefully just code for dealing with errors
- possibly run-tagging also, but likely hold off on kar redesign
- discussion of semantic search / workflow search and browse / aaron's work
- aaron: component search should still work fine
- lsids and search need to work for 2.0
- discussion of tagging
- working pretty well now, inserts tags into moml like semantic tagging
- sean will write summary summarizing what's done and left to do, so we can figure out where to draw the line for 2.0
- distributed computing discussion
- is master slave ready for 2.0? jainwu: needs another use case. if it works with that, then yes
- hadoop - working on using hadoop as alternative kepler distributed computing option
- next week:
- TODO aaron, derik, sean: synopsis of your work and what needs to be done for 2.0
- TODO daniel: updates to eng view page, send email for if we should discuss this on next call, or special call
- TODO jones: discuss hadoop etc w/ jianwu
|
