[od-discuss] compatibility should (was Re: Status of Vancouver and Surrey OGL varients)

Fri May 15 19:24:56 UTC 2015

On 12/10/2014 12:18 AM, Mike Linksvayer wrote:
> On 12/09/2014 11:52 PM, Paul Norman wrote:
>> On 11/20/2014 1:45 PM, Mike Linksvayer wrote:
>>>
>>>      A bigger issue to me is that I've been told by governments using
>>>      them that the OGL-BC derived licenses (OGL-Surrey, OGL-Vancouver)
>>>      are not compatible with CC BY, ODC-BY or the ODbL. Because the
>>>      Canadian OGL variants are essentially only usable by a single
>>>      government, this leaves me with an impossible situation for making
>>>      works from multiple sources and my work.
>>>
>>>
>>> That's also extremely annoying but wouldn't make the licenses non-open
>>> per the definition.
>>>
>> While I agree in principle that an open license need not be compatible
>> with other open licenses, I am worried by licenses which are
>> interoperable with neither CC or ODC licenses.
>>
>> I am not sure that a non-reusable license which fails to be compatible
>> with any other open licenses can be considered open. If I take Vancouver
>> data, modify it and add my own copyrightable contributions, and want to
>> release the new work, I cannot do so under any open license. Other open
>> licenses are not compatible, and I cannot release it under the
>> OGL-Vancouver license as I am not the City of Vancouver.* To my mind,
>> this is non-open.
>
> I'm very sympathetic to your argument, but we (whoever was participating
> at time anyway, I haven't dug up discussion) decided to not address
> compatibility in OD 2.0. The fallback was to require that a license be
> compatible with at least one of CC-BY-SA, ODbL, or GPL in order to be
> put in "recommended" category.
>
> I've added https://github.com/okfn/opendefinition/issues/77 for 2.1 to
> make sure this gets re-evaluated

I've created a pull request addressing this, rationale below for 
discussion, copied from https://github.com/okfn/opendefinition/pull/103

I continue to think compatibility an important issue and that the 
definition is by far the most powerful venue this group has, so should 
(heh) be mentioned there.

We have other *should*s in the definition, this one is as important.

I've kept this very brief:

 > The **license** *should* be compatible with other open licenses.

Some obvious ways to extend:

 > The **license** *should* be compatible with other open licenses to 
the maximum extent possible given its policy ends.

or

 > The **license** *should* be compatible with other open licenses, in 
particular one or more of the most commonly used copyleft licenses.

However I tend to think generality is good here, message should not 
prompt thinking of excuses (which ...policy ends... does) nor narrowness 
(which ...copyleft licenses does).

Mike