I'd start by looking through https://incubator.apache.org/guides/graduation.html and seeing if you're in ready. Assuming you conclude you're ready, the next step is to start a [DISCUSS] thread on dev to see if others in the community agree and to start building the graduation proposal to send to the IPMC and eventually the board.
Alan. On Thu, Jun 14, 2018 at 3:59 PM Andreas Neumann <a...@apache.org> wrote: > Yes I agree, it is time to approach graduation. What would be the first > steps? > > On Thu, Jun 14, 2018 at 9:49 AM, Alan Gates <alanfga...@gmail.com> wrote: > > > As long as Tephra has a group of PMC members willing to carry it > forward, I > > think graduation is a good path. I'm not trying to push you into being a > > subproject. > > > > Andreas, do you agree that Tephra is ready to start thinking about > > graduation? > > > > Alan. > > > > On Thu, Jun 14, 2018 at 8:48 AM Andreas Neumann <a...@apache.org> wrote: > > > > > My 2 cents: > > > > > > I think in its current scope, Tephra is quite mature and complete. > > > > > > The next big step would be to implement a fully distributed transaction > > > manager, with replication and consensus protocol to ensure consistency. > > > That would be a major undertaking and I am not sure that it would be > > > feasible as a sub-project of another project that uses HBase. I would > be > > > concerned that it could turn into a utility/library specifically for > that > > > other > > > project. > > > > > > If Tephra were to join another community, I would suggest HBase itself > > > to keep it generic and widely usable. > > > > > > -Andreas > > > > > > > > > > > > > > > On Tue, Jun 12, 2018 at 8:20 PM, Alan Gates <alanfga...@gmail.com> > > wrote: > > > > > > > If you have a viable community of developers who are interested in > > > ongoing > > > > development (even if it is at a slower pace) then I think you're > ready > > to > > > > start moving towards graduation. Having a project that focuses on a > > > niche > > > > case isn't a problem, as long as you have the developers to keep > moving > > > it > > > > forward. If you are only being used in one case, Phoenix, it might > > make > > > > sense to approach that community and see about joining them. But it > > > sounds > > > > like that isn't the case. > > > > > > > > Alan. > > > > > > > > On Tue, Jun 12, 2018 at 5:35 PM Poorna Chandra <poo...@apache.org> > > > wrote: > > > > > > > > > Hi Alan, > > > > > > > > > > Tephra is now mostly feature complete for the basic transaction > > > use-case. > > > > > We need to now start working on features like replication, > > scalability, > > > > > etc. When we start working on those features, you will again start > > > seeing > > > > > activity in terms of JIRAs and patches. > > > > > > > > > > Transactions on HBase is a still a pretty niche segment. I'm not > sure > > > how > > > > > widespread the adoption is going to be. Although the scale of the > > > > > developmental activity is small, I feel Tephra has met most most of > > the > > > > > requirements for graduation. What would you recommend in such a > case? > > > > > > > > > > > > > > > Poorna. > > > > > > > > > > On Thu, Jun 7, 2018 at 10:07 AM, Alan Gates <alanfga...@gmail.com> > > > > wrote: > > > > > > > > > > > It seems development on Tephra has slowed quite a bit lately, > with > > > > only 7 > > > > > > JIRAs filed during the last 3 months. I know Tephra is used in > > other > > > > > > projects, like Apache Phoenix. Is this slowdown temporary? Is > > > > > development > > > > > > slowing because Tephra is mature and only maintenance is now > > > required? > > > > > Is > > > > > > it continuing to be picked up by new projects or only used in 1 > or > > 2 > > > > > > places? I ask because it feels like it is near time for Tephra > to > > > > > graduate > > > > > > (it's been incubating for over 2 years now) and what needs to be > > done > > > > to > > > > > > graduate is somewhat determined by the answers to these > questions. > > > > > > > > > > > > Alan. > > > > > > > > > > > > > > > > > > > > >
