Avik, Sorry for the delay in replying to your questions. See my replies inline. I know we also talked about some of this while in person at the Mentor Summit.
On Sun, Oct 1, 2017 at 8:57 AM, Avik Ganguly <avikganguly...@gmail.com> wrote: > Hello everyone, > > 1) Context :- Last time I merged my organization's MifosX fork, I faced > difficulties in deciding whether to rebase with MifosX or Fineract. On > going through the commit history of both in details, I found MifosX had > more assets related changes whereas Fineract had more liability and > infrastructure related changes. (I went with Fineract in the end as the > only thing we really wanted from MifosX was rescheduling which had open > issues being addressed by Pramod at that time). Based on the commit > history, I am still surprised that one set of client apps is able to > support both platforms (which means API contracts has not diverged yet). > Question :- Am I missing out on something or are the efforts of the > community ending up in 2 similar codebases for some particular reason? > Avik - after we entered incubation at Apache, the mifosx repository/codebase was deprecated. As you can see from the repo at https://github.com/openMF/mifosx, the last commit to that codebase was in February 2016. There shouldn't be any differences in terms of asset or liability functionalities in Apache Fineract all the code (apart from Pentaho reports libraries and Hibernate-related code) was moved into Apache Fineract when mifosx repository was deprecated. As I think you've already discovered, the main repo for Apache Fineract is at https://github.com/apache/fineract/. We simply maintain a mirror of it from openMF github. All pull requests and forking of the repo should be from the apache/fineract repo. All that being said, Mifos X the distribute/community release on top of Apache Fineract maintained by the Mifos Initiative is still active. But it merely consists of the .WAR file bundled with the web app that is built on top of the latest stable release of Apache Fineract. The blog post we made a while back tries to explain the infrastructure of version control, communication, and issue tracking. I welcome any updates to that page or incorporating this into the Apache Fineract wiki itself: http://mifos.org/blog/infrastructure-apache-mifosx/ > 2) Is there any volunteering or dedicated QA engineer who certifies > releases other than Santosh? Jenkins making PRs breaking existing tests > un-mergeable was part of MifosX; any reason for it not being part of > Fineract? > > Right now for Apache Fineract 1.0 and the Mifos X distribution and apps, Santosh on contract from Conflux Technologies leads the role. However, over the longterm and especially with Fineract-CN this has to be a community-led effort by members of the PMC, individual partners, and volunteers. Let's start a separate thread on getting continuous integration successfully running for Apache Fineract. I'm not sure if there are any restrictions from the Apache side. > 3) Any particular reason why the release process is mentioned as a > bottleneck? Is cherry-picking MifosX commits or MifosX picking Fineract > commits somehow involved as part of release process? Is Nazeer contributing > to both MifosX and Fineract releases? > What is most time-consuming (at least in the past) has been the back and forth to get code up to community standards and get it generic enough for shipping with the main release. Nazeer is the lead who guides both the Apache Fineract release at the Apache level and he also guides the packaging of the subsequent Mifos X distribution release on top of Apache Fineract. > > Regards, > Avik. > > On Sat, Sep 16, 2017 at 9:33 AM, vishwas <vish...@confluxtechnologies.com> > wrote: > > > Thanks Myrle ! > > > > Regards, > > Vishwas > > > > > > > > > On Sep 15, 2017, at 12:20 AM, Myrle Krantz <my...@apache.org> wrote: > > > > > > Ed and vishwas, > > > > > > I've filed an Infra ticket to request moderation privileges for both of > > you. > > > > > > Here: > > > https://issues.apache.org/jira/browse/INFRA-15089 > > > > > > Moderation is done via a klugy email interface. Instructions on how > > > to use ezmlm to moderate once you have those privileges can be found > > > here: > > > https://reference.apache.org/pmc/ml > > > > > > Regards, > > > Myrle > > > > > > > > > On Thu, Sep 14, 2017 at 6:43 PM, vishwas > > > <vish...@confluxtechnologies.com> wrote: > > >> Hello All, > > >> > > >>>>> 1.) Are you doing something that only you are doing, and that you > > >>>>> believe should be shared more widely? > > >> > > >> No > > >> > > >>>>> 2.) What are the reasons why you aren't doing one of the activities > > on > > >>>>> our "bottleneck" list? > > >> > > >> I am currently focusing on a few other things I really want to do and > > would not be able to take up time consuming activity like reviewing pull > > requests. > > >> > > >> I would be happy to help with moderating lists though. > > >> > > >> Regards, > > >> Vishwas > > >> > > >> > > >> > > >>> On Aug 31, 2017, at 9:05 AM, Gaurav Saini <gauravsain...@gmail.com> > > wrote: > > >>> > > >>> Hello Myrle, > > >>> > > >>> Sorry for delayed reply. Below are my answers > > >>> > > >>> On Tue, Aug 29, 2017 at 6:57 PM, Myrle Krantz <my...@apache.org> > > wrote: > > >>> > > >>>> Hi Fineracters, > > >>>> > > >>>> There are 9 PMC members who have not responded (8 if we excuse > > >>>> Nazeer). As a reminder, the questions are: > > >>>> > > >>>> 1.) Are you doing something that only you are doing, and that you > > >>>> believe should be shared more widely? > > >>>> > > >>> > > >>> No > > >>> > > >>> > > >>>> 2.) What are the reasons why you aren't merging PRs or helping out > > >>>> with releases? > > >>>> > > >>>> > > >>> My main expertise was on frontend mainly, but recently after looking > in > > >>> Fineract CN I am really interested to get started with Fineract CN. > > >>> > > >>> Also, I saw that new Fineract CN has a frontend project based on > > >>> typescript, material so I am really to interested to contribute on > that > > >>> right now also. > > >>> > > >>> Thanks > > >>> Gaurav > > >>> > > >>> > > >>>> I would appreciate a response. You might also want to say, "I'm too > > >>>> busy right now", or "I'm waiting for Fineract CN", or "I've lost > > >>>> confidence that the project can help me achieve my goals". If > that's > > >>>> the case, then it is important information too. > > >>>> > > >>>> Thanks, > > >>>> Myrle Krantz > > >>>> > > >>>> On Thu, Aug 17, 2017 at 10:06 AM, Myrle Krantz <my...@apache.org> > > wrote: > > >>>>> Hi all, > > >>>>> > > >>>>> We have an awesome community. In particular, we've been very good > at > > >>>>> responding to people's request for help getting Fineract working > for > > >>>>> their use cases. But we also have some things we need to work on. > > >>>>> > > >>>>> Specifically we have several bottlenecks in our community: > > >>>>> > > >>>>> 1.) Only Nazeer is merging PR's. > > >>>>> 2.) Only Nazeer is serving as a release manager. > > >>>>> 3.) Only Myrle is moderating the lists (apart from the mentors who > > >>>>> will eventually step out). > > >>>>> > > >>>>> This is called "having a truck factor of one", because if a truck > > hits > > >>>>> one of us (or if we win the lottery and decide never to do anything > > >>>>> for Fineract again), then the entire project would fail (1). This > is > > >>>>> already causing us problems in some areas: for example our PR's are > > >>>>> backed up > > >>>>> > > >>>>> I'm not writing this to lecture anyone. I know there are reasons > for > > >>>>> the way things are. That doesn't mean we should just accept it > > >>>>> though. > > >>>>> > > >>>>> Let's start by improving our understanding of the problem. Would > > each > > >>>>> committer please say: > > >>>>> 1.) Are you doing something that only you are doing, and that you > > >>>>> believe should be shared more widely? > > >>>>> 2.) What are the reasons why you aren't doing one of the activities > > on > > >>>>> our "bottleneck" list? > > >>>>> > > >>>>> > > >>>>> I'll start: > > >>>>> 1.) List moderation. > > >>>>> 2.) I'm not merging PR's because I'm not sufficiently familiar with > > >>>>> the current generation Fineract code to feel comfortable evaluating > > >>>>> those PR's. I'm not serving as a release manager, because I don't > > >>>>> have enough insight into when we are release-ready, and because > there > > >>>>> are several pieces of technical knowledge I'd need to acquire to > be a > > >>>>> release manager which I don't yet have. For these reasons, I've > > >>>>> chosen to focus my energies on Fineract CN instead. > > >>>>> > > >>>>> Best Regards, > > >>>>> Myrle Krantz > > >>>>> > > >>>>> P.S. After we have a list of causes, I'll initiate a round of > > >>>>> discussion about how we can resolve the causes. > > >>>>> > > >>>>> 1.) http://www.agileadvice.com/2005/05/15/agilemanagement/ > > truck-factor/ > > >>>> > > >>> > > >>> > > >>> > > >>> -- > > >>> Regards, > > >>> *Gaurav Saini* > > >> > > > > > -- *Ed Cable* President/CEO, Mifos Initiative edca...@mifos.org | Skype: edcable | Mobile: +1.484.477.8649 *Collectively Creating a World of 3 Billion Maries | *http://mifos.org <http://facebook.com/mifos> <http://www.twitter.com/mifos>
