A major milepost is a "release" that people should use instead of trunk? Should Mahout push for that amount of fit&finish? Keep plowing ahead on new algorithms? Keep plowing ahead on tools to handle new problems?
Lance On Sat, Mar 26, 2011 at 6:51 AM, Robin Anil <[email protected]> wrote: > On Fri, Mar 25, 2011 at 8:42 PM, Sean Owen <[email protected]> wrote: > >> Maybe it's the first flush of springtime that's reinvigorated activity, but >> I do see significantly more activity in the last week or two. That's great. >> Especially since a lot of the activity is going into older issues too. The >> number of open issues is declining; it was over 70 a month or two ago and >> is >> now down to just over 50. That's a good trend. >> >> Our 0.4 release was Oct 31 2010. Given a general pattern of releasing every >> 6 months, that would suggest end of April for 0.5. I see no reason to rush, >> but reason to start asking those release questions: >> > +1, Faster releases matter a lot > >> >> - What do we want out of 0.5 as a release? I suggest it be viewed as nearly >> a release candidate for 1.0; the APIs and functionality should be >> substantially set by 0.5, to be polished for a 1.0 release in Q4. >> > Lets focus first on getting the progress we made since 0.4 out. Polishing > can continue in the summers > > >> - What is marked for 0.5 that just isn't realistically going to be done by >> anyone in a few weeks? > > - What isn't in JIRA for 0.5 that would be nice for 0.5? > > Let me take a shot at scrubbing issues as well. > -- Lance Norskog [email protected]
