Hey, Some remarks/questions inline.
What are your thoughts on this? Summary; * Aggregate a la DDD.
I think it'd be a good opiniated move.
* Event Store model.
If I understand correctly this means that EventSourcing would be built in Qi4j persistence support. In other words, we would gain EventSourcing "for free". Am I right?
* Transition events in store.
I see TimeSeries coming :)
* Asynch model for store/retrieve.
Would it mean that underlying stores "transactions" (or whatever it is) should be done in a separate thread?
Finally, is this too much of a change for 2.0, and should be scheduled for a 3.0 right away?? *I* think this is the right way to go, and I think some areas in the persistence system will be simplified, as well as setting the stage for "historic data" support, stronger distribution capabilities, event sourcing model and much more...
I wouldn't bother scheduling this for 2.0, it depends on the amount of work it is. On one hand it could delay 2.0 and impact motivation but on the other hand theses changes are really interesting and would break compatibility so they can't be in a 2.x release. I have no strong opinion on this yet but I have the gut feeling that it will have impacts that need discussion (what about query?, migration? etc..). To be continued. /Paul _______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
