We use github for ravendb and require CLA for anything except tests We have a LOT of people contributing
On Tuesday, January 15, 2013, Stefan Bodewig wrote: > On 2013-01-13, Troy Howard wrote: > > > One thing to consider is that the ASF workflow is a bit stodgy to work > > with. It really does make a difference to collaboration, as any small > > barrier makes it too easy to opt out of contributing and collaborating. > > OK, I bite ;-) > > What you describe (and I have largely snipped now) is how github pull > requests lower the barrier to entry and how "send a patch and attach it > to a JIRA ticket" is more cumbersome. > > Over at log4net we have people struggle with creating patches. What > otherwise seem to be competent developers fail to use diff or even use > svn diff. Using git would be even more difficult for them. This may > not be the target of "get more people involved in development", I'd > aggree with that. > > I can't contest that pull requests are really easy for people who have > done that a few times. But there is a bit more to it. > > In my limited experience with pull requests it seems they create more > "drive-by contributions" than the more heavy weight process at the ASF. > The fact that forking and creating pull requests is so easy seems to > result in pull requests by people that never show up again after it has > been merged. It feels as if people who have jumped through the hoops to > actually contribute are more prone to stick around. I may be wrong. > > Another aspect is that many times forks are created to fully implement a > feature in isolation that somebody feels is worthy. If you start out by > proposing the feature and discussing it with the existing community > first you may realize the majority doesn't want the feature or that a > completely different way to skin that cat is more appropriate. I don't > say you couldn't have those discussions up front in a "pull request > world", it just doesn't seem to happen as often. Again, I may be wrong. > > The github "pull request world" seems centered around the code itself > much more than the traditional way of the ASF which believes to be more > centered around the community of developers. > > I'm not saying the "pull request world" is inferior (nor it is superior) > but rather that at least I'm not sure how well the "pull request" view > mixes with the ASF philosophical goals. > > Then again, we are free to use git and I'd be all for trying out how > well it works for us. After all a pull request at github really just is > some icing on top of what you can do with the git CLI anyway. Nothing > would stop you from manually merging a fork into your local git repo and > push the change to the ASF repo - you just didn't have the button inside > a Web-UI. But this difference is only visible to the existing committer > base, not to the developers who want to contribute. > > Stefan >
