Hi Do Yung, Thank you very much for the clear explanation. I've recently applied for a mentor for this project, so I'm reading past messages to understand the context now. ;-) Please see my comments inline.
On 2018/10/31 06:56:15, DO YUNG YOON <sho...@gmail.com> wrote: > It's been almost 3 years since s2graph has entered apache incubating. Our > community has not been successful to reach more people so far and I think > failing on growing community is the blocker issue for s2graph to graduate. > > I want to open up the discussion about what we have to do for growing > community, and eventually for graduation. > > In my opinion, there are many parts have been changed after incubation. > > Followings are notable improvements we have been working on. > > 1. Query Interface > > - Tinkerpop OLTP support > - GraphQL support > > 2. Storage > > - Embedded RocksDB > - JDBC Storage > > 3. Full-text search > > - ElasticSearch for search on vertex/edge property > > 4. OLAP > > - S2GraphSource: dump HFile then creates Spark dataframe. > - S2GraphSink: sink Spark dataframe to S2Graph. > > Even though the project has been improved, our community has not been > grown. > > There could be many reasons that our community is not growing(please list > them so we can discuss if we can resolve them), but in my humble opinion, > we have not even released above improvements since the last release. > > As far as I remember, the last time we discuss the release was a long time > ago(2017-09-08) at here > https://lists.apache.org/thread.html/d0d8e264b7625128a6c2832723cf2ad555d169f9f277086ce69c3a7b@%3Cdev.s2graph.apache.org%3E > . > > Back then I was under pressure that we need to show that our community > growing after release, and since I was not sure, I avoid talking about the > release, and rather keep working on issues. > > Now I think it is much better to start by releasing it rather than only > just wondering how to grow the community. > > To grow the community, the very basic way I know is to keep improving the > project, releasing it, and advertise it, and get user feedback, and I think > we stuck to step 1 for too long time. My two cents: >From my experiences, it's really about selling your technology to people. We >might want to be focused on marketing a little bit more. Possible marketing >activities for community growth are in my mind: - Write blogs with compelling stories for potential active users who need to live on this technology, potentially with runnable GitHub project(s) for readers Stories with real use cases attract people as their lives can rely on those. It could make them active users at least. - Make more noise through Kakao, Twitter, Facebook, ... I have been a non-social-web guy, but I have written blog articles for Apache Jackrabbit, etc. JMeter twitts often to broadcast community activities. It works. - If necessary or useful, consider writing those in Korean for local communities, even in users@ mailing list. As long as you keep deciding together in devs@, it's no problem. - Try to leverage your company's marketing department's power. They might be eager to leverage you as an Apache community leaders for their reasons, too. It could be a win-win. For example, I always try to secure my tech session(s) about Apache Jackrabbit and related topics in my company's annual conference, making deals with the marketing dept. ;-) > > After a few more releases, then if there is no one found s2graph useful, > then I think we can happily retire from apache incubating, but I just don't > feel like we did our best yet. +1. I see many good things in this project with great potentials. > > I am willing to pursue building community, and I think it is worth to try > out a few more release until decide what's the next step. > > I want to ask what others think. are there any others who are willing to > get more involved? I'm learning now. I hope to get involved more soon. :-) Kind regards, Woonsan >