Dev Meeting Thursday Sept 18.

1. 0.2 Meeting this afternoon. Anyone who wants to make changes to the source tree should attend; it seems this is only Ducky and Chao. Looks like 0.2 will be tomorrow, and then we'll move on to 0.3.

2. 0.3 -- This will be a calendar release as well. In other words, we determine the date and release what we have. The 0.3 release will be in January sometime; we'll figure out what to do about Chrismas as we get closer.

3. Discussion about what people think they want to get done for 0.3.

• John wants to make progress on the document architecure, get rid of the viewer parcel and do a range of other related things (sorry, I missed part of this discussion).

• There is a chunk of repository work to be done that is important to everyone. There will be a separate meeting for the prioritizaton of Andi's tasks for the repository.

• Andy wants to get agents working. Andy's not planning on doing a lot of work on the contacts parcel for 0.3. He thinks contacts should be rewritten using the document architecture, and we should wait till that's a little but further along before plunging in too deeply.

• Jed will be working on the document architecture.

• Brian has more issues about the data model that would be good to resolve. We'll figure out what else Brian will be working on.

• Morgan is thinking about start-up performance, build size and installers.

• Mitchell will be working with the wxWindows project a bit, since that project is important to Chandler. The other big goal is to get back to community activities: what should be done to encourage and assist community development in the 0.3 timeframe, and what resources are necessary to make things happen.

• Ducky will be working with Chao on feature prioritization, and has offered to work on QA.

• Michael Toy as virtual Katie (as Katie is away would like to get to a calendar which is usable or daily work. We call this "dogfood" s in "we need to eat our own dogfood" or the "dogfood calendar." It means something that is not necessary ready for the general public, but which is usable by the intrepid. But the infrastructure isn't ready yet; so spending a lot of time on the calendar itself might not make sense just yet. Katie might help John in the document architecture world; we need to decide.

• Michael has a contractor is looking at interaction of Python and Mozilla mail libraries to see if this is a useful path. Michael wants to hire a good chunk of people, though this is stressful for the organization. Michael would like 0.3 to look better than 0.1 and 0.2. We expect to have a graphical person on board soon.

• Mitch plans to get focused in the design area, pick some areas , define deliverables and deliver them. He also wants to figure out how to support others in the organization, and to make sure that our 0.3 activities are part of a longer -term community plan.

• John for Robin: switch to modern version.

• Heikki wants to figure out what wxWindows work is important for Chandler, and where that is in the queue. He also plans to look at high level security architecture.

• Chao will be focusing on feature prioritization

• There are also some holes in our 0.3 planning -- areas where people aren't yet working.

4. Andy raised the topic of when releases won't be based on a calendar, but on features, quality and polish. For example, 0.2 is happening halfway through the repository conversion, so there is more than 1 repository in the release. It's possible that 0.3 will fall halfway through the document architecture. When do we plan a release based on what it actually does? On the other hand, there is a discipline to calendar-based releases. A suggestion was made to do the development of 0.3 on a calendar. Then review where we are, and see what needs to be done to make it a good, clean release. No decision on this, more thinking needed.

5. Discussion of development process for 0.3. Should we continue the 2 week build cycle we had for 0.2? People seemed to like this, and we should continue this method for now. Brian notes that it would be good to improve our learning on accurate scheduling. General agreement on this point.

-- MitchellBaker - 20 Sep 2003