Q: What conferences should OSAF be sending people to next year? Thinking of any presentations?
α4
stability is the goal, will hold release for bug fixes
current release date: 17 october
anything we should be doing differently to discover, diagnose or fix bugs?
α5
EIM, sharing format, dump and reload
calconnect
Notes
Q: What conferences should OSAF be sending people to next year? Presentations?
Andi: wants to go to europycon in July
Grant: IETF might not provide much benefit for OSAF, unless we have some sort of agenda. Calconnect is good for interop testing.
Heikki: has been wanting to go to PyCon
Morgen: quite happy not to attend any conferences
Bear: happy to attend conferences on the east coast
PJE: PyCon would be nice, main things he might present related to 20% project and setuptools. Networking opportunities will be useful.
α4
Morgen: Prefetching items for performance might be destabilizing -- include in α4?
Heikki: should try and fix performance regressions
Andi: can we just get the thing to work first? stability is more important.
Katie: stability bugs should have higher priority. Morgen's work on performance won't be a waste of time, though. Even if we decide not to accept a fix for α4 if the risk isn't worth it, we'll take it 2 weeks later.
α5
PJE sent out his implementation plan for EIM. He should have something ready for Morgen to get started in a few weeks.
Dump and reload needs to be spec'd further, not worked out so not clear if it will fit in α5. PJE's layer will be implemented, but will require work at a higher layer that is not spec'd and doesn't have an owner. Should build on top of sharing, so somewhat sequential.
Calconnect
4 servers: Cosmo/Apple/RPI/Oracle. Oracle changes our timezones into UTC, which causes confusion. Apple server worked perfectly. They have a pretty big test script, don't have a lot in the way of regressions? Mikeal is looking at making use of it.