Help needed: Multiverse 0.3 and datastructures
Peter Veentjer
I have checked in the 0.3 stuff. A lot of previous limitations are
removed and calls can be done on 'normal' objects (so no handles
needed anymore).
One of the things I'm looking for is people willing to create
datastructures that can participate in STM. That is why I have
created the org.multiverse.datastructures package that contains stuff
like queues, stacks etc. But this package needs a lot of improvements:
- work with interfaces
- more datastructures (skiplists, trees, hashtables etc)
Essentially it is a rewrite of java.util and java.util.concurrent
classes but tailored for STM. So who wants to be the next Joshua
Bloch/Doug Lea :)
So I'm looking for volunteers that want to pick this up. You don't
need to dive in the depths of the stm implementation, because these
datastructures should be stm implementation independent. You can
focus on writing datastructures that can participate in the STM and
providing feedback on what needs to be improved in Multiverse.
I'm hoping that more than 1 person can pick this up so we can help
each other out.