On Sat, Oct 31, 2015, at 03:29 PM, Arvind Prabhakar wrote: > Thanks for the explanation David. > > The concern that Joe and I raised several months back is about where > > project direction and other decisions are being set. > > > > > > Work was being done, reviews happening in public, but there didn't > > seem to be any discussion on future direction of the project, new > > features, new integrations. They'd show up as bugs in Jira, and then > > code started appearing. > > > While I understand the concerns described in your mail, I feel that these > are rooted in your expectations and not necessarily a problem with the > community/project. I have, for example, worked on projects where a large > number of ideas and discussions started on JIRA and not the mailing > lists... Sqoop, Oozie, Flume, and even Hadoop projects are thriving > examples of that. Sentry leans that way too, and I see nothing wrong with > that approach. Note that discussions on JIRA do get back to the mailing > list, albeit in a different format than a direct email. > > Projects like these typically have ad-hoc/organic evolution. They do not > necessarily have a master plan for setting the direction beyond the basic > idea of their mission. As long as the project is seeing development and > adding value to it's user community, I think it is doing it's job well. > > When Sravya and I were talking in Budapest, she indicated that there > > were a number of Hangouts/phone calls happening where things were > > being discussed, and potentially even being decided. I didn't know > > these were taking place. I asked that at a minimum that folks bring > > back discussions to the mailing list. (See email from end of > > September) I also suggested that maybe instead of 2-5 people going off > > and setting up a hangout or conference call that the project have it's > > own call, known to everyone, with open invites so that others could > > attend, and meeting notes kept. > > (N > Lots of projects have local meetups that don't get talked about on the > mailing list. I agree that those should be advertised on the mailing list > and care must be taken to ensure those who are interested can attend. I > feel the same way for hangouts and phone calls. I feel that the Sentry > community is committed to doing this as openly as possible and bringing > back the discussions to the mailing lists, thanks to your nudging on > this. > > All that aside, I do want to go back to my original question for you: is > there an objective criteria that this project must meet before it can > graduate? I would love to have that outlined crisply so that I can do my > job as a mentor to help them get there.
Not speaking for David, but for myself: - Preferably, discussions would happen on the mailing list and enable people not already involved in Sentry development to become involved and participate. - Decisions would not be taken off list, period. - If the project wishes to do its work in Hangouts or phone meetings, then it needs to advertise those ahead of time and make it possible for others to participate. Notes from those meetings need to make it back to the mailing list and it needs to be possible for others to participate equally if they can't make real-time meetings. (Note, I don't feel what I've bulleted out is dramatically different from what David is saying. I'm not sure how his examples are non-objective.) Best, jzb -- Joe Brockmeier [email protected] Twitter: @jzb http://www.dissociatedpress.net/
