What were the road-blocks from moving the issues to github again? On Fri, Apr 29, 2011 at 2:38 PM, Aaron S. Meurer <[email protected]> wrote:
> Does anyone have any ideas on ways that we can better manage the pull > requests at GitHub and the issues at Google Code? Right now there are some > issues: > > - People forget to reference the issue number on the pull request or the > pull request on the issue. The best case is that I notice this and fix it, > but it's possible that people don't even realize that there is an issue for > the pull request or visa versa. > > - Even when people remember to reference the pull request on the issue, > either they forget or don't have permissions to add the NeedsReview and > username labels. I think this can be partially fixed by making sure that > all regular contributors are members in the Google Code project so that they > can add labels to issues. I used to be able to handle this manually pretty > easily, but lately the traffic has increased by quite a bit (because of the > GSoC students) and it's getting harder to manage. Ironically, this makes > the labels even more essential. > > - People forget to check the corresponding issue before pushing in a pull > request, like as happened with issue 2302/pull 257. I am just using this as > a recent example, I am not trying to blame Chris here. > > - It's easy to type the wrong issue or pull number. Pull 257 is also an > example of this (and again, I don't want to blame anyone here; the point is > that typos are easy to make). > > Does anyone know if it would be possible to have some kind of more > automated system, where most or all of the above are done automatically? In > particular, it would be nice if the following could be automated: > > - If a pull request is created, then the issue should be marked as > NeedsReview with a label of the person's name. This could be based on > referencing an issue number in a pull request. > > - If someone puts the NeedsBetterPatch or even PassedReview label on an > issue, the pull request should be notified. > > - All pull requests should have an issue, I think. Maybe something could > warn someone if they forgot to reference an issue number in a pull request, > and ask them if they want to create a new issue for it, or reference an > already existing one. > > - In the pull requests, automatically change "issue 1234" to "[issue 1234: > solve(4**x-x**4, x) (and similar) should give all solutions]( > http://code.google.com/p/sympy/issues/detail?id=1234)", or add that as a > new comment. This would make it so that you could see the issue title, > which would let you verify that the number was correct, and it also would > auto-link it, which would make navigation easier. > > With that, you would just have to make sure that you reference the issue > number in the pull request (it could also parse commit messages), and the > rest would be taken care of. > > I'm thinking this should be done by some kind of bot somewhere, but even a > script that you run from your computer and goes through and it does it from > your account (after checking to see if it has already been done or not) > would work. > > I'm putting out a request to the community, because I don't know how to > program such a thing. Does this sound like a good idea? Does it sound > feasible? > > Aaron Meurer > > -- > You received this message because you are subscribed to the Google Groups > "sympy" group. > To post to this group, send email to [email protected]. > To unsubscribe from this group, send email to > [email protected]. > For more options, visit this group at > http://groups.google.com/group/sympy?hl=en. > > -- You received this message because you are subscribed to the Google Groups "sympy" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/sympy?hl=en.
