Think about a pull request as a patch, that is the easier way to think about it. And yes, we would probably treat them in the same manner as we do patches.
I would assume that the most common behavior would be that each of the team members would clone from the master repository, so everyone clones are local. Whenever we push, we push the the mainline. There are multiple CI solutions for GIT, it is pretty easy to handle from that respect. Code review remains the same, every member of the team is responsible for reviewing any code that they push to the main line. On Sun, Jan 24, 2010 at 6:59 PM, Fabio Maulo <[email protected]> wrote: > a pull-request is about what ? IMO each pull request should be related with > a JIRA if it is related to a new feature or a bug-fix. > More then that I would understand > - which is the responsibility of each member about "his own" fork and the > main-line. > - The CI > - who will have the responsibility of code-review > and so on > I would like to read something more because I'm pretty sure that "anything > will work as today but more quickly" is not the true. > > 2010/1/24 Ayende Rahien <[email protected]> > >> The general behavior would be pretty much the same. >> We already own github.com/nhibernate, so that can be the master >> repository. >> Only committers are going to have write access there. >> Users will be able to do one of the following, either submit a pull >> request from their fork, or submit a patch (like they do today). >> I don't see it affecting the JIRA in any way. >> >> On Sun, Jan 24, 2010 at 6:45 PM, Fabio Maulo <[email protected]>wrote: >> >>> btw Oren I don't have a problem to move on Git if we can talk about the >>> organization of the team and what should be the work-flow to push patches >>> coming from users (for instance who will write the JIRA to have a track >>> and/or if we can push something without a test). >>> >>> >>> 2010/1/24 Ayende Rahien <[email protected]> >>> >>>> By anyone who reads the code. >>>> I am trying to commit my changes to SVN now (3rd time that it failed), >>>> but here are the commits as they were done on the git version (attached as >>>> patch files). >>>> Using git, I could make as many commits locally as I want to, and then >>>> push them up to the master repository in a single action. The granularity >>>> in >>>> which we could work suddenly becomes much easier. >>>> >>>> On Sun, Jan 24, 2010 at 6:35 PM, Fabio Maulo <[email protected]>wrote: >>>> >>>>> I know... and you have used the word is scaring me "easy to review"... >>>>> review by who ? >>>>> >>>>> 2010/1/24 Ayende Rahien <[email protected]> >>>>> >>>>> It isn't just that, it is the fact that I can do commits locally. >>>>>> Just to give you an idea, the local work I did for the lazy props >>>>>> feature has taken about ~17 commits when done on git. >>>>>> Each of those was a specific change, with a comment and easy to >>>>>> review. >>>>>> With SVN, I had to do 1 commit, which is much harder to handle. >>>>>> >>>>>> On Sat, Jan 23, 2010 at 5:26 PM, Richard Brown (gmail) < >>>>>> [email protected]> wrote: >>>>>> >>>>>>> I haven't used Git (other than to pull the FluentNHibernate sources >>>>>>> to my machine) ... but the ability to get a history of a file when >>>>>>> disconnected (on the train) sounds useful to me, so I'm game for >>>>>>> learning it >>>>>>> if we move. >>>>>>> >>>>>>> >>>>>>> *From:* Fabio Maulo <[email protected]> >>>>>>> *Sent:* Saturday, January 23, 2010 1:15 PM >>>>>>> *To:* [email protected] >>>>>>> *Subject:* Re: [nhibernate-development] Subversion is killing me >>>>>>> >>>>>>> btw Oren, If I'm recalling right there are some NH-forks in Git-Hub, >>>>>>> perhaps you can use some of those forks and when the work is done you >>>>>>> can >>>>>>> commit the done-work in the SVN. >>>>>>> >>>>>>> 2010/1/23 Fabio Maulo <[email protected]> >>>>>>> >>>>>>>> You can't do it in the trunk ? >>>>>>>> >>>>>>>> >>>>>>>> 2010/1/23 Ayende Rahien <[email protected]> >>>>>>>> >>>>>>>>> It is slow. >>>>>>>>> Case in point, I wanted to do the lazy prop branch, commit and >>>>>>>>> switch to the new branch. >>>>>>>>> I an't do that, it keeps failing. >>>>>>>>> >>>>>>>>> >>>>>>>>> On Sat, Jan 23, 2010 at 2:31 PM, Fabio Maulo <[email protected] >>>>>>>>> > wrote: >>>>>>>>> >>>>>>>>>> I don't have problems with SF SVN. >>>>>>>>>> I have had some problems in the past but related with some SF >>>>>>>>>> reorganization, now everything work fine. >>>>>>>>>> >>>>>>>>>> Which problems are you experimenting ? >>>>>>>>>> >>>>>>>>>> 2010/1/23 Ayende Rahien <[email protected]> >>>>>>>>>> >>>>>>>>>> I just spent hours trying to get things working right with SF >>>>>>>>>>> SVN. >>>>>>>>>>> It is slow, it fails a lot and when it does, it completely breaks >>>>>>>>>>> my ability to work. >>>>>>>>>>> >>>>>>>>>>> I don't think that it had been discussed, but I want to bring it >>>>>>>>>>> up, can we switch to Git for the project? >>>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> -- >>>>>>>>>> Fabio Maulo >>>>>>>>>> >>>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> -- >>>>>>>> Fabio Maulo >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Fabio Maulo >>>>>>> >>>>>>> >>>>>> >>>>> >>>>> >>>>> -- >>>>> Fabio Maulo >>>>> >>>>> >>>> >>> >>> >>> -- >>> Fabio Maulo >>> >>> >> > > > -- > Fabio Maulo > >
