www-infrastructure-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Roy T. Fielding" <field...@gbiv.com>
Subject Re: Labs, was Re: New lab, using Git
Date Fri, 18 Jun 2010 20:06:52 GMT
On Jun 18, 2010, at 11:58 AM, Paul Querna wrote:

> On Fri, Jun 18, 2010 at 10:36 AM, Roy T. Fielding <fielding@gbiv.com> wrote:
>> On Jun 18, 2010, at 9:41 AM, Philip M. Gollucci wrote:
>>> Hash: SHA1
>>> On 06/18/10 12:19, Jukka Zitting wrote:
>>>> Hi,
>>>> Since there were no objections so far, I'll move forward on this over
>>>> the next few days.
>>>> I'll use labs@ for topics related to the proposed Oak lab itself, and
>>>> infra-dev@ for stuff related to the required Gerrit setup.
>>>> BR,
>>> Jukka, sorry, I've not been paying enough attention due to real life.
>>> <Infra VP Hat>
>>> There are several issues [1] with what you propose.  Some of them we can
>>> fix others we can't.  Please hold off for now until we discuss this some
>>> more.  I think at this point a better venue would be infra-private until
>>> we bring the results back to the public eye.
>>> 1.  You're running a production service, and of all services Version
>>>       control in a ZONE.
>>> 2.  Nobody knows how this works other then you.
>>> 3.  Just b/c its a lab doesn't mean its free of the ASF wide policy
>>>       guidelines.
>>> </Infra VP Hat>
>> On the contrary, because it is a lab it is not able to do releases
>> and is not an Apache project.  It is therefore outside any of our
>> policies regarding Apache releases, including use of subversion.
>> I don't think there should be any more restrictions placed on labs
>> than we would place on what committers do with their home directory
>> on people.a.o.
> The only thing we have consistently done for committer home directories is:
>  1) Don't have security issues (ie, world writable files)
>  2) Don't use of excessive space or resources.
> We don't require CLAs for content in committers home directories.
> There is no required license of content on
> http://people.apache.org/~username.

There is no required license of content in Labs either, though
it is implied in most cases.  It isn't necessary given the CLA.

> When a projects is in Labs, even though there are not releases, we
> assume the person committing to it is licensing it to the ASF just
> like any other project under their CLAs -- and saying that they don't
> need to adhere to the same policies seems out of place.

They do need to adhere to Labs policies and they are restricted
to existing committers.  They are not allowed to do releases.

> Labs projects have previously been accepted to the Incubator without
> any additional CLAs or other attempts to verify their IP, mostly
> because previously they have been held to the same policies as
> existing ASF projects in all other matters.

No, it is because their content comes from a committer with
a CLA on file.  That is all that any contribution needs.

> Saying that they do not need to adhere to ASF policies because they
> don't have releases seems like a fundamental shift in what the Labs
> PMC was presumed to be doing, and at the very least other PMCs like
> the Incubator must understand this, if we really believe that Labs
> should only have the same restrictions as we have on home directories.

They don't need to adhere to release policies because they don't
do releases.  Incubation is frigging irrelevant to content that
has been contributed under CLA and contributing to a lab is
certainly covered under that agreement.  It makes no difference
whether that contribution comes in the form of Git or Subversion or
Jira or Bugzilla or Mail or written down on a gigantic role of
toilet paper and thrown in the general direction of ApacheCon.

Please, we have enough policies already.  If we need any more then
they will be imposed by the Labs PMC, not by infrastructure.
Infra only needs to step in if our assets are being misused or
there is some harm being caused to other projects.

View raw message