Hi Andreas,

The push request is an interesting feature of github.

However, Apache requires some rules, especially, if an user provides a patch, he should grant the "ASF donation". It's the only way that we are sure that the user want really donate the code.
With a push request, we don't have this step.

That's why checking the box in Jira is important. It's legal requirement, especially around IP.

Maybe you saw some IP clearance thread. For instance, to migrate a maven plugin from codehaus to apache, we create a Jira containing the whole patch, just to be able to check the "ASF grant" and have a clean IP clearance.

Even if it's true that, from a technical point of view, github push request is really great, from an Apache legal point of view, I'm not sure it's applicable.

I'm gonna talk with other members to have their point of view.

Regards
JB

On 08/19/2011 06:06 AM, Andreas Pieber wrote:
Hey guys,

Just because it is annoying me a little bit the last days I would like to
hear your opinion about this before nagging the infrastructure team if this
is possible at all. The scenario is the following:

While I've absolutely no problem with the current ASF ways (provide patches
via JIRA&  central SVN) I think it is quite a pain to either review code
and/or discuss ideas&  "code-snippets". On the other hand, with github&  git
getting bigger and bigger, ASF already with mirrors on github and github
with a REALLY great code review functionality I think the following feature
could be useful:

if a pull-request is added to our fork at github the typical notification
mail is not send to a unique person but rather to our dev list. If we reply
on this mail it will be (a) mirrored to the github discussion thread and (b)
(as the identity of the dev-list I assume) back to the github discussion
thread. If we (or the provider of the request) add a comment to the
discussion thread the notification will be sent to the dev list (same
thread) again. Since this can create some amount of traffic I think this can
also be a [email protected] list. Once we conclude that a
patch-request from an user is good enough the patch could be created and
attached to an JIRA.

IMHO this would make the entire code-review and dicussion process ways
easier and still public accessible and followable by all commiters/devs.

WDYT?

Kind regards,
Andreas


--
Jean-Baptiste Onofré
[email protected]
http://blog.nanthrax.net
Talend - http://www.talend.com

Reply via email to