Ubuntu Linux - a productized Debian. Perhaps we should be looking at this in the future.
Notes
Project management / organization / status
The Apps group is now using Wiki notebook pages for reporting status and keeping daily work records. I slowed down on my wiki notebook pages because its a pain to create the pages and link them in. I wonder if this usage would be better served by giving each person a blog -- the blog software would take care of linking/archiving/etc. We could use the same software that the ASF is using for PlanetApache? to display a group blog, which would aggregate all the posts into a single place. It would also have the benefit of providing a full RSS feed, which would make it easier to read all the status, etc.
The Basecamp link (above) shows come good examples of the kind of functionality that I'd like to see in a project dashboard.
If you name your pages intelligently, you don't need to link them in. See SmallMeetingNotes, for example -- the Community pages are all linked in via a query. One disadvantage of a blog is that it's not as well set up for making changes at a later date. While in theory, the Journal is for things that don't change over time, in reality, many people go back (usually within about a week) and make changes. -- DuckySherwood
0.5 Tasks
Developer Dogfood
help develop plan
implement items (unknown)
pylucene plan
front page, web log, etc.
increase use of transparent communication mid-process rather than only at the end
Performance
trace tool
co-ordinate with Andi - work on requirements
TBD
Query Language
Query Grammar
better notation for recursive kind queries
better notation for kind queries -- omit WHERE TRUE
handle floating point literals
bug with spaces in strings
clean up parameter syntax
clean up handling of path literals? (depends on addressing)
allow use of Pythonisms like infix 'is', 'contains'
yapps patches to Amit
Full Text
allow restriction to particular collections/kinds
Query semantics
queries over stamped (multiple superkinded) items
queries vs defaultValue/initialValue
allow parameters in paths?
orderBy (sorting) support - interaction with ref-collections
expand set of functions that can be used in queries
interruptible queries (generators vs ??)
queries vs item clouds -- do we want a way to preload a cloud using query notation?
Documentation
Overview
link to relevant docs/api docs/specs
spec
samples
Query Processor
Notification
better notification for path queries
implement notification using monitors - change in semantics, must co-ordinate with ItemCollection?
implement notification for full text queries
Full Text
support for full text indexing at the parcel.xml level (co-ordinate with Morgen)
Python API
factory methods for query creation (eliminate need to pass repository)
queries are items and query results are ref-collections