> Another question on this thread, are you asking everyone to switch to this > version for running all of Mahout, if so wouldn't this entail first > re-running a unit tests on this beta version and certify that there are no > issues. Let us know more details. > > Not quite; I just wanted to gauge the general level of interest in this sort of undertaking, i.e. if anyone else has run into problems that would be significantly easier to solve with certain aspects of the newer APIs. If interest was widespread enough, then yes we could have that conversation.
I'd suggest we not require something later than a latest stable > release at this phase, if there are other workarounds. > Another option is to copy in the code to Mahout for now. I don't know > if that's feasible. > I do have other options I can fall back on; it will simply take a lot more coding. Trade-offs either way :)
