|
Nov 09, 2009
Participants: Barseghian, Crawl, Jones, Ludäscher, Riddle, Schultz, Wang
- Barseghian
- Parviez in india, will try Eric for tagging/wrm feedback.
- tags and errors now (de)serializing properly in runs, now importing errors and tags into provenance
- still some bugs with tagging and exporting/importing kars.
- TODO: after a call that Riddle will setup to discuss tagging/WRM UIs among REAP/Kepler people, setup a meeting with scientists (Eric, Jim, Oliver, et al) for their perspective
- TODO: schedule Wang devlunch
- Jones
- tried tagging, had a number of problems with saving and opening. Also UI needs discussion.
- Schultz
- working on getting away from moml parsing to speed things up
- TODO: consolidate and update bugs
- Wang
- Master-Slave doesn't work w/ new build system, will fix this
- Hadoop and mapreduce work also needs to be changed to work w/ new build system. target? post-2.0. Hadoop needs testing on bigger cluster, more functionality needs to be added
- TODO: send email w/ mapreduce and master-slave bugs
- TODO: Dec16 devlunch discussing distributed computing
- Altintas
- Master-slave before mapreduce. Mapreduce work will be ongoing, let's release that when it's more complete in a couple months
- discussion of migration scripts that will be needed for schema changes.
- TODO Altintas or Crawl file a bug for tracking 1.0 => 2.0 hsql schema update
- Crawl
- Working on Engineering View Interfaces
- Looking at Metacat scheduler code
- Riddle
- Tagging more than one thing can be really slow, refactoring
- TODO: setup a call to discuss usability with reap/kepler people.
- Probably skip next week's call - Matt, Peter, Ilkay away.
|
