Hey all- A data point and observation from an ASF Member and Incubator PMC Member...
Moon is correct that readiness for graduation is a function of community development and adherence to the Apache Way, rather any specific feature or tech milestones. Since entering Incubator, Zeppelin's had two relatively easy releases, has finished the incubation checklist (http://incubator.apache.org/projects/zeppelin.html), has added new commiters, etc. In short, Zeppelin's in a good position to graduate from my perspective. Resolution of specific PRs should be handled in a speedy matter, but there doesn't seem to be any disagreement to that - just some work left to be done in getting them in. -Jakob On 3 February 2016 at 23:39, Amos B. Elberg <amos.elb...@gmail.com> wrote: > I don't see a point to splitting it. The reason we didn't merge in December > is that bugs in CI prevented the tests *in* 208 from functioning. It wasn't > causing anything else to fail. Now CI is broken for the project anyway. If it > was going to be split, I would do that myself. > > The reliability of the code has been proven in the field: People who don't > use R have switched to the version of 208 in my repo because it compiles > reliably when 0.5.6 does not. > > This has been outstanding since August, and it's very hard to understand a > reason - you even participated in a Meetup in September where a variant of > the code in the PR was used as a demonstration of Zeppelin's capabilities and > potential. > > Part of being an Apache project is dealing with significant PRs from outside > the core development team. Addressing these issues for R, and Prasad's PR, > seems like a good test of project maturity. These were features on the > roadmap which were supposed to be included before the first non-beta release. > >> On Feb 4, 2016, at 12:50 AM, moon soo Lee <m...@apache.org> wrote: >> >> Thanks Sourav for interest in this discussion and very valuable opinion. >> >> I completely agree how much R and Authentication (which i believe already >> in Zeppelin) will be useful for users. And believe me, I want these >> features in Zeppelin more than anyone. >> >> But at the same time we have diversity of user bases. >> Some people might think supporting general JDBC is more practical and more >> useful feature, the other can think multi-tenancy is the most important, >> etc, etc. >> >> So, i believe Apache Top Level project is defined by how community works, >> not defined by what feature does the software includes. >> >> Regarding bringing R into main branch, I tried to make pr208 passes the CI. >> I could able to make it pass 1 test profile, but couldn't make it pass all >> other test profiles. >> >> I'm suggesting split the contribution into smaller peaces and merge one by >> one. Like Hayssam did it for his contribution of Shiro security integration >> (pr586). And I'm volunteering making pr208 into smaller PRs. >> >> Best, >> moon >> >> On Thu, Feb 4, 2016 at 2:11 PM Sourav Mazumder <sourav.mazumde...@gmail.com> >> wrote: >> >>> This does make sense Moon. Completely agree with you that features are not >>> important for becoming a top level project >>> >>> However, in my opinion, from the practical usage standpoint, without these >>> two features Zeppelin does not look to me a full fledged top level project. >>> Curious whether there are any technical glitches which are impediment in >>> bringing these features to the main branch. Wondering if any help can help >>> to get those problems fixed. >>> >>> Regards, >>> Sourav >>> >>>> On Wed, Feb 3, 2016 at 4:22 PM, moon soo Lee <m...@apache.org> wrote: >>>> >>>> Hi guys, >>>> >>>> I don't think any feature (R or whatever) should be prerequisites of >>>> graduation. Especially when a project never setup those features as a >>>> graduation goal. >>>> >>>> Including specific features could be valid concern for release >>> discussion, >>>> but i don't think it's related to a graduation. >>>> >>>> Graduation is much more like if project is doing it in apache way, in my >>>> understanding. Last time the reason why i didn't go for a graduation vote >>>> is, because of there were valid concern about contribution impasse. >>>> >>>> Since that, community improved / clarified contribution guide and review >>>> process. And Zeppelin PPMC members were trying to help many contributions >>>> that they have been as a open PR for a long time. (Especially Jongyoul >>> and >>>> Felix helped a lot) >>>> >>>> So, let's move discussions like 'which feature should be included' to the >>>> release / roadmap discussion. >>>> In the graduation discussion, i'd like to have an discussions, such as >>>> evaluating >>> http://community.apache.org/apache-way/apache-project-maturity-model.html, >>>> etc. >>>> >>>> Does this make sense for you guys? Amos, Eran, Sourav? >>>> >>>> Thanks, >>>> moon >>>> >>>> On Thu, Feb 4, 2016 at 8:05 AM Amos B. Elberg <amos.elb...@gmail.com> >>>> wrote: >>>> >>>>> No Eran is right. The last vote for graduation passed-it was not >>>> withdrawn >>>>> in favor of releasing 0.5.6. It passed and there was some feedback from >>>> the >>>>> mentors concerning graduation, R, and some other issues. And that's the >>>>> last public discussion about graduation until today. >>>>> >>>>> Alex if you disagree with that do you have links to the discussion >>> emails >>>>> that you're referring to. >>>>> >>>>>> On Feb 3, 2016, at 2:34 PM, Alexander Bezzubov <b...@apache.org> >>> wrote: >>>>>> >>>>>> Hi Eran, >>>>>> >>>>>> thanks for sharing your oppinion! >>>>>> >>>>>> Could you please check my previous reply about release schedulle and >>>> let >>>>> us >>>>>> know if that makes sense to you? >>>>>> >>>>>> By the way, please our mentors correct me if I'm wrong here, but >>> after >>>>>> reading [1] I was under impression that project does not have >>>> pre-request >>>>>> regarding its code or features in order to undergo this formal >>>> procedure >>>>> of >>>>>> graduation. >>>>>> >>>>>> [1] >>> http://incubator.apache.org/incubation/Incubation_Policy.html#Graduating+from+the+Incubator >>>>>> >>>>>>> On Wed, Feb 3, 2016, 20:04 Eran Witkon <eranwit...@gmail.com> >>> wrote: >>>>>>> >>>>>>> If I recall correctly R support was one of the pre-requisite for >>>>> graduation >>>>>>> from day one. >>>>>>> I agree that Authentication should be added as well. >>>>>>> +1 for graduation after we add both >>>>>>> Eran >>>>>>> On Wed, 3 Feb 2016 at 20:24 Sourav Mazumder < >>>>> sourav.mazumde...@gmail.com> >>>>>>> wrote: >>>>>>> >>>>>>>> Surely I vote for the same. Zeppelin is already very popular in >>>>> different >>>>>>>> quarts of the Spark/Big Data user group. High time to graduate it >>> to >>>>> top >>>>>>>> level. >>>>>>>> >>>>>>>> However, I shall suggest to have the support for R and >>> Authentication >>>>>>> added >>>>>>>> to Zeppelin before that. These are the supports most of the people >>>> are >>>>>>>> eagerly waiting for. >>>>>>>> >>>>>>>> Regards, >>>>>>>> Sourav >>>>>>>> >>>>>>>>> On Wed, Feb 3, 2016 at 8:23 AM, moon soo Lee <m...@apache.org> >>>> wrote: >>>>>>>>> >>>>>>>>> Thanks Alexander for resuming the discussion. >>>>>>>>> Let's start a vote. >>>>>>>>> >>>>>>>>> Best, >>>>>>>>> moon >>>>>>>>> >>>>>>>>>> On Wed, Feb 3, 2016 at 11:11 PM Alexander Bezzubov < >>> b...@apache.org >>>>> >>>>>>>>> wrote: >>>>>>>>> >>>>>>>>>> Dear Zeppelin developers, >>>>>>>>>> >>>>>>>>>> now, after number of releases and committers grew more I'd like >>> to >>>>>>>>> suggest >>>>>>>>>> the re-new the discussion of graduating Zeppelin to top level >>>>>>> project. >>>>>>>>>> >>>>>>>>>> If there are on objections - next step would be to start a VOTE >>>>>>> thread >>>>>>>>>> here. >>>>>>>>>> >>>>>>>>>> What do you guys think? >>>>>>>>>> >>>>>>>>>> -- >>>>>>>>>> Alex >>>
