On Sep 15, 2016 8:31 AM, "Alex Harui" <aha...@adobe.com> wrote: > > > > On 9/15/16, 12:34 AM, "Christofer Dutz" <christofer.d...@c-ware.de> wrote: > > >Just to add my 50ct to this discussion. > > > > > >Justin did bring up the issue. He even brought it when we were first > >discussing starting a release. The discussion sort of dried out without a > >resolution, then when it came to the release, he mentioned it again. But > >again no action was taken. So I too see a complaint about PMCs coming up > >with such stuff in the last minute as not valid. > > > > > >ASF is community over code, but the one thing the ASF deals with is > >protecting us as developers as well as our users from licensing issues. > >It's this extra protection and care that distinguishes Apache releases > >from the typical Github projects. This is why in every bank or insurance > >company I was working in, there never was a discussion about using ASF > >software, if it's not ASF however you have to jump quite a lot of > >obstacles in order to use a library. I remember quite some rounds with > >legal and quality assurance people. > > > > > >The ASF has earned that trust because we have people in our communities > >that care about this sort of stuff. Having some legal caretaker is one of > >the coolest thing a project can have, cause it lets us coding-monkeys do > >what we like to do and we can somehow be lazy and trust that someone is > >taking care of this. I hate legal stuff. To me it's just getting in my > >way ... sort of as others think working on build-, code-quality or > >writing documentation should be done by others. > > > > > >We have a pretty heterogenous community. I know I'm definitely the > >build-monkey, Justin's the legal-monkey, we have a lot of code-monkeys. > >Why not use the specialities of each other instead of complaining about > >it? I know I have to work on my side about not ranting about > >code-quality, for example, but I'm trying ... hope you guys didn't notice > >any recent rants from my side ;-) > > That's exactly it, Chris. We do appreciate Justin helping make sure that > the big-ticket items regarding licensing are in order, and we appreciate > you trying to improve code-quality, but if your "itch" is a higher > priority to you than to others, you have to choose the right words and > time to encourage the community to change. Looks like you are wise enough > not to try to get us to stop and resolve code-quality issues just before > the release. > > But, code quality is more objective in that there are tools that say you > got it right or not, and you can scratch that itch yourself without > impacting others. This isn't true for many licensing issues. The answer > isn't always clear, and often the message feels more like Justin is > telling someone else to do the work instead proposing "hey, I think we > should apply this patch to this file". > > And again, we have to consider community energy and time vs the ASF's > reputation. For all of these recent issues agh we did not resolve, > pushing their resolution off to the next release will not harm the ASF's > reputation but has a good chance of helping build the community. > > > > > > >If we hadn't let the discussion about Justin's findings die when he > >brought it up and had resolved the problem instead, the problem would > >have been solved. So how about us addressing the issues Justin has and in > >case of a "I think this way, you think that way", let's involve legal and > >have these things settled once and for all? > > > > This is a puzzle I still haven't solved. Folks like Justin, who are > spread across multiple ASF projects and have limited time, may not be able > to keep the discussion going in a timely fashion. If you look at the > recent issues, I generally respond to Justin's comments, then the clock > starts ticking. How long should we wait for his response? Silence often > means consent. I think it is better for the community for us to wait a > little bit, but I can see when Justin is active on other Incubator issues > so I know he isn't off-the-grid, so then we should just keep on going. > Then the question becomes: when Justin does return to the discussion and > is not satisfied but it is now late in the game, what is the best way for > him to respond? If it were me, I would say "hey, I'm still not happy with > this, so let's revisit in the next release" and not vote at all, but it is > true that he is certainly free to keep voting -1 on our releases. > > -Alex >
Lisencing issues, no matter when they come up should be tracked in JIRA like every other issue. And just like every other issue, we can always fix it in a forthcoming release. I would leave it up to the Release Manager to make the call on a case by case basis, whether to hold the release or to punt it to the next release. Thanks, Om