On Tue, May 15, 2012 at 10:02 PM, Tibor Mlynarik <[email protected]> wrote:
> How is [1] and [2] related to current org.qi4j.library.eventsourcing library ? The current event-sourcing is not an optimized event store and it is not intrinsic to Qi4j Core. I would like to see that the event store is a first class citizen just like the Entity Stores are, and that an SPI exist for making those pluggable and that they are part of a UnitOfWork. > Last time I needed to propagate UnitOfWork.resume/pause calls to the > entitystore SPI, without nice solution. > Is mentioned EventBus going to be used internally by Qi4j runtime to > allow such extensions/hooks into qi4j internals ? I don't understand the question. I have not analyzed the requirements for an Event Bus yet, but I see evidence in other programs that I write that it is a very handy abstraction, which in Qi4j could be even nicer since we have Visibility and Module/Layer/App scopes, Dep Injection and a way to make things pluggable when needed. > To comment on [3], can you point me where is multiple-type problem discussed ? It was on this mailing list, just 2-3 months ago when I expanded the multi-type scope in Qi4j. Previously, a Composite had one base type, and that is going away, and it will be possible to have many types on the Composite of equal weight. > Is it just entity store persistence/query problem or is it broader > scope problem ? The issue is roughly like this; * Previously, all data stored was typically contained as a single JSON object for all Mixins state of a Composite Type. If I asked for a TwoLegged composite, the type resolution would return a Human and the state loaded would contain the entire Human. If Ostrich was also visible, it would fail in the type resolution. * Now, if I ask for TwoLegged, it will not go "via" the Human but simply load the Identity under the TwoLegged type, and the state loaded should only be that of the supertype, yet the remaining state must not be lost. (harder to explain...) > Is EntityComposite going to lose it's main type ( currently first in > CompositeDescriptor.types() ) ? Correct. Cheers -- Niclas Hedhman, Software Developer http://www.qi4j.org - New Energy for Java I live here; http://tinyurl.com/3xugrbk I work here; http://tinyurl.com/6a2pl4j I relax here; http://tinyurl.com/2cgsug _______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
