[foundation-board] CKAN licence?

Thu Aug 25 10:28:57 UTC 2011

On Thu, Aug 25, 2011 at 10:59 AM, Becky Hogge <becky.hogge at gmail.com> wrote:
> On 25 August 2011 10:46, Ben Laurie <ben at links.org> wrote:
>> On Wed, Aug 24, 2011 at 6:47 PM, Rufus Pollock <rufus.pollock at okfn.org> wrote:
>>> On 24 August 2011 15:55, Ben Laurie <ben at links.org> wrote:
>>>> On Wed, Aug 24, 2011 at 3:04 PM, James Casbon <casbon at gmail.com> wrote:
>>>>> Its AGPL it's in the README, I was handicapped by trying to use my phone.
>>>>
>>>> Well. That leads to two questions:
>>>>
>>>> a) Which README? Not the one in the source repo.
>>>
>>> It was in the README but got I see it was accidentally removed by
>>> person doing a docs refactor a couple of weeks ago. Now back:
>>> <https://bitbucket.org/okfn/ckan>
>>
>> I am not too concerned about this, but my understanding is that if you
>> want a licence to apply properly to source you need to mention it in
>> each source file.
>
> That strikes me as concerning. Jordan, can you advise?
>>
>>> (A bit confusingly we maintain a github mirror which is not up to date
>>> and hence contains the licnese text in its README:
>>> https://github.com/okfn/ckan).
>>>
>>>> b) Why the AGPL? Yes, I realise this could be a long conversation, so
>>>> let me open with my extreme preference for the Apache Licence. Why?
>>>> Because to me "open" means "you can use it for any purpose". The AGPL
>>>> is not a licence that supports that. There's a longer argument which
>>>> is about barriers to adoption and the value of non-coerced
>>>> contribution.
>>>
>>> Open in relation to us means compliant with Open Source Definition
>>> which the AGPL does.
>>
>> This is hardly an answer, as the Apache Licence also does.
>>
>>> There is a much bigger discussion, which you have
>>> begun on, around share-alike clauses which I suggest taking up when we
>>> meet in person :-)
>>
>> Happy to do so, but that may be some time since I am in Wales for the
>> next n months. As I suggested before, a phone call might be a good
>> idea in the interim.
>>
>> In any case, though, is this not a matter for the whole board?
>
> Or perhaps wider - the Board aspire to limit itself to matters of
> legal and financial compliance, and to conduct discussions about
> strategy in a way which meaningfully includes the wider community.

Whilst I broadly agree with this goal, licensing discussions are hard
to have with a wide community because you very quickly rathole into
the GPL vs BSD wars, which seem unresolvable to me (it seems patently
obvious to me that BSD style licences, amongst which I count the
Apache Licence, are more open than GPL style licences - but apparently
the opposite is patently obvious to others). In licensing it seems
there's a need for some amount of benevolent dictatorship. Or at least
a benevolent oligarchy.

Another interesting question is: what is the wider community? Allowing
a free-for-all, as the IETF does, means that people with time on their
hands get to strongly influence the discussion, even when they have no
stake whatsoever in the outcome. It seems to me that community should
be limited to active participants. However, participation is somewhat
governed by licensing - you won't find GPL zealots working on Apache
projects, and vice versa. So asking a community who have chosen to
contribute to a GPL-based effort whether it should be GPL-licensed is
likely to be a foregone conclusion.

In any case, this discussion is important to me, whatever the forum,
and I am not in favour of delaying it.