Hi everyone, I'm the other half of the team that did the generifying (if
that's even a word) of collections. Sorry I've been silent so far during
the discussion of what to do with the new collections, I was out of town
and just got caught up with the debate.
So after reading the discussion and seeing it stall out a bit, I propose
that we start by creating a new project in apache which will contain the
current source code from our collections.sf.net repository. To this new
project we can all then add things like new MultiMap and
ChainedTransformer interfaces and implementations that will break
compatibility with existing Collections releases. We can then also
accept new generics specific test cases to add to the already existing
set. The specifics of how all this would work, i.e. if John and I would
become apache committers or not, I'm not sure what the best route here
is. If this is to become a semi-autonomous project then I think having
us as the main committers probably does make the most sense.
Throughout the process I think it's important to maintain a beta or
alpha release of the project so early adopters can start using it and
contributing feedback. There is no advantage in my mind to holding off
on releases until we have the last few controversial classes converted,
better to get feedback from the entire community on the bulk of the work
up front.
I think with our working implementation of the majority of classes,
along with the collection15 team's input and contributions on the
remainders we should be in great shape overall.
Matt
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
