Mailing-List: contact legal-discuss-help@apache.org; run by ezmlm
Precedence: bulk
Reply-To: legal-discuss@apache.org
Content-Type: text/plain; charset=us-ascii
Mime-Version: 1.0 (Mac OS X Mail 10.2 \(3259\))
Subject: Re: When is a CCLA/ICLA required for a contributor?
From: Jim Jagielski <jim@jaguNET.com>
In-Reply-To: <0c5a98baad83403ab06a1b8e91f87895@gsdgamp04etn3.firmwide.corp.gs.com>
Date: Thu, 23 Mar 2017 08:33:42 -0400
Cc: "Maxin, John" <John.Maxin@gs.com>
Content-Transfer-Encoding: quoted-printable
Message-Id: <EC305134-C430-411C-B6D0-110313B76CF9@jaguNET.com>
References: <0c5a98baad83403ab06a1b8e91f87895@gsdgamp04etn3.firmwide.corp.gs.com>
To: legal-discuss@apache.org
archived-at: Thu, 23 Mar 2017 12:37:04 -0000

For the ASF, an iCLA is required once one gets commit access.

Regarding the CCLA, that must be determined my the individual =
themselves.
There is no one, complete, all-inclusive answer other than "it depends" =
:)

> On Mar 22, 2017, at 5:06 PM, Rezaei, Mohammad A. =
<Mohammad.Rezaei@gs.com> wrote:
>=20
> Apologies if the answer to the question is obvious. After reading the
> documentation on apache.org and searching through this mailing list, I
> could not reach a definitive conclusion. I also apologize for the=20
> length of this inquiry. I had a hard time making it shorter.
>=20
> Definitions: I'm only interested in a contributor, not a committer.=20
> In my understanding, a committer is someone who has commit rights to =
an
> Apache hosted/owned repo. Everyone else who would like to affect a=20
> change in an Apache hosted/owned repo is a contributor.
>=20
> I realize there are two questions here:=20
> For a contributor, when is an ICLA required?=20
> If an ICLA is required, when is a CCLA required?
>=20
> For the sake of brevity, I'm going to assume that if a contributor is=20=

> required to sign an ICLA, a CCLA would be required if they are =
employed=20
> under plain US law and the work is in any way relatable to their=20
> employer's business. I'd like to concentrate on the first question=20
> (ICLA) from this point on.
>=20
> I'm basing my analysis on the following:
>=20
> (A) on this page: http://www.apache.org/licenses/
> the statement "The ASF desires that all contributors of ideas, code,=20=

> or documentation to any Apache projects complete, sign, and submit=20
> (via fax or email) an Individual Contributor License Agreement (ICLA)"
>=20
> (B) In the thread "code without an iCLA ..."
> http://markmail.org/message/lv6vcnfqvygk2xv3
> The consensus appears to be "ICLAs are required for everything=20
> that is non-trivial." (as explained by Sam Ruby).=20
>=20
> (C) On this page: http://www.apache.org/dev/contributors.html
> there is no mention of any ICLA requirement.=20
>=20
> (D) In the thread "Re: Requiring CLAs for all contributions"
> http://markmail.org/message/cnbxal5mheof2hmy
> Mark Thomas (mar...@apache.org) said:
> "I see lots of downsides to that policy and no upsides.=20
> The issues that concern me:=20
> - Unnecessary barriers to entry for new contributors=20
> - Legally unnecessary=20
> - Creates additional work for the secretary (not much now but if every
>    TLP and podling did this it might)=20
> - Re-enforces the meme an iCLA is required to contribute at Apache=20
>=20
> I'm also don't believe that an iCLA is even desired for all=20
> contributions. The ALv2 provides all the legal cover we need and =
desire."
>=20
> (E) In the thread "Re: FAQ: CCLA Optional?"
> http://markmail.org/message/5ch6kvovjow32esk
> Henri Yandell (bay...@apache.org) said:
> "Committers sign ICLAs, contributors don't (typically - maybe a =
project
> has a reason for requiring contributors to sign ICLAs but I don't =
recall
> it being done)."
>=20
> (F) on this page: http://apache.org/foundation/how-it-works/legal.html
> 5 possible paths for incoming code is outlined.
>        1) "An existing third party project joining Apache"
>              is about projects joining ASF, so not relevant
>        2) "A large one off code contribution ... One off code =
donations
>              can come in through a software grant".=20
>              appears to open the door for not requiring an ICLA,=20
>              but a software grant.
>        3) "Repeated contributions applied directly to the source": for=20=

>              committers, so not relevant.
>        4) "Patches contributed via the issue trackers"
>              seems to have no ICLA requirement. No size limit=20
>              is discussed.
>        5) "Patches contributed via mailing lists are expected=20
>              to be simple."=20
>              No ICLA requirement, but an unspecified size limit.
>=20
> (G) It also appears that different Apache projects interpret this=20
>    differently:
>    =
http://flink.apache.org/how-to-contribute.html#submit-a-contributor-licens=
e-agreement
>      requires an ICLA -- always.
>=20
>    =
https://beam.apache.org/contribute/contribution-guide/#potentially-submit-=
contributor-license-agreement
>      "We require you to have an ICLA on file with the Apache Secretary=20=

>      for larger contributions only."
>=20
>=20
> The 3 possible answers for the question of ICLA requirement are =
analyzed
> below:
>=20
> 1) It's always required.
>   (A) is strong evidence that ICLA is always required. There is no=20
>   procedure outlined for a contributor to opt-out of ASF's "desire",=20=

>   rendering it a requirement.
>=20
>   in (B), the distinction between trivial and non-trivial requires
>   advanced degrees in law, philosophy, and information theory; =
therefore,
>   any mere mortal should just sign the ICLA.
>=20
> 2) It depends...
>   In (A) the word "desire" is not the same as "require". "Desire" =
signals
>   a possibility to opt-out on the part of the contributor.
>   (B), the determining factor being "non-trivial". Parts of (F-5) =
support this.
>   (G) It may also depend on the particular project.
>=20
> 3) It's not required.
>   (C), assuming such an important matter would not simply be left off.
>   (D),(E)
>   (F-4), so long as all the work is contributed via JIRA
>=20
> Can a clear, simple answer be given to this question? I don't consider
> "It depends" to be clear or simple, unless the rules could be enforced=20=

> by a small piece of computer code with no room for interpretation.
>=20
> Thanks
> Moh Rezaei
>=20
>=20
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: legal-discuss-unsubscribe@apache.org
> For additional commands, e-mail: legal-discuss-help@apache.org
>=20


---------------------------------------------------------------------
To unsubscribe, e-mail: legal-discuss-unsubscribe@apache.org
For additional commands, e-mail: legal-discuss-help@apache.org