Hi, >the number of >different Microkernel implementations is definitely low and the current >way of acquiring them via the MicroKernelFactory seems over engineered.
I don't think this is over engineered. But I think the MicroKernelFactory class has too much implementation specific code now. There are use cases that benefit from a mechanism to acquire a MicroKernel instance in a "not hardcoded" way, for example the IndexWrapper implementation, and all other wrapper implementations. But it would also help for testing and so on. >I suggest to remove the MicroKernelFactory and use standard constructor >calls as you proposed for now. That would break quite a lot of existing code and use cases. > Further down the line we should probably >rethink the issue of modularisation, dependency injection and >configuration within a broader scope anyway which might make this issue >obsolete. See also OAK-17. Yes we can do that, but we should do that before we remove the MicroKernelFactory. Regards, Thomas
