[ZF] Repository choice: Please indicate your preference until Sunday 9/23

[Chris McDonough]

On 09/18/2012 12:03 PM, Martijn Faassen wrote:
> Hi there,
>
> I prefer #2, github only. If not #2, then #1, and finally #3.
>
> A distributed version control system offers some complexity to the
> requirement that write access is only by signed contributor. This
> because someone who hasn't signed a contributor agreement can start a
> fork of a project and then do a pull request. Someone who has signed
> the agreement can then merge the fork. But the code was in fact
> contributed by someone who hasn't signed any agreement. This pattern
> is a very common way people contribute to git projects, especially on
> github which makes forking and pull requests very easy.
>
> In the past we've had manual patches of this nature, but these are
> generally quite small so there are less copyright concerns.
>
> So are we going to require anyone who offers a pull request has signed
> the contributor agreement before it is merged? If so, how are people
> easily going to find out that someone offering a pull request has
> signed said agreement? If this is hard, that'll retard merges or might
> result in merges that haven't been validated.

I don't think this really changes anything.  Legal documents have never 
required every person who has supplied a patch to sign a contributor 
agreement in the past.  It has always been at the discretion of each 
committer to merge a patch he's been provided, and he has not been bound 
to create a trail of where that patch originated from.  Merging pull 
requests instead of applying emailed patches won't change this at all, 
except possibly it will be more transparent.

- C