Re: contrib/fuzzystrmatch/dmetaphone.c license - Mailing list pgsql-hackers

From Michael Paquier
Subject Re: contrib/fuzzystrmatch/dmetaphone.c license
Date
Msg-id CAB7nPqSxkWvWYVxNu4sjN2drDoDMtQzDoq3+rs87fp16QdiTEw@mail.gmail.com
Whole thread Raw
In response to Re: contrib/fuzzystrmatch/dmetaphone.c license  (Stephen Frost <sfrost@snowman.net>)
List pgsql-hackers
On Thu, Feb 26, 2015 at 8:56 AM, Stephen Frost <sfrost@snowman.net> wrote:
> * Jim Nasby (Jim.Nasby@BlueTreble.com) wrote:
>> On 2/25/15 4:10 PM, Andrew Dunstan wrote:
>> >
>> >On 02/25/2015 11:59 AM, Joe Conway wrote:
>> >>>
>> >>>It's largely because of such uncertainties that I have been advised
>> >>>in the past (by those with appropriate letters after their names)
>> >>>to stop using the Artistic licence. This is why I spent nearly a
>> >>>year working on changing pgAdmin to the PostgreSQL licence.
>> >>I committed this (1 July 2004), but cannot remember any details about
>> >>a license discussion. And I searched the list archives and curiously
>> >>cannot find any email at all about it either. Maybe Andrew remembers
>> >>something.
>> >>
>> >>I doubt we want to rip it out without some suitable replacement -- do we?
>> >>
>> >>
>> >
>> >That's more than 10 years ago. I remember creating this for my then work
>> >at the North Carolina State Highway Patrol and sending it to Joe, but
>> >that's about the extent of my recollection.
>> >
>> >If the Artistic License isn't acceptable. I guess we'd have to try to
>> >get the code relicensed, or reimplement the function ourselves. There
>> >are numerous implementations out there we could copy from or use as a
>> >basis for reimplementation, including several licensed under the Apache
>> >2.0 license - is that compatible with ours?
>>
>> Perhaps a company large enough to have in-house counsel
>> (EnterpriseDB?) could get a quick legal opinion on the license
>> before we start pursuing other things? Perhaps this is just a
>> non-issue...
>
> For my 2c (IANAL), I'm not convinced that it's an issue either..  I've
> certainly not heard of anyone complaining about it either, so..
>
> That said, we could also through SPI which might get us a bit of
> pro-bono work, if we really wanted to pursue this.  Just a hunch, but as
> they tend to be conservative (lawyers in general), I expect the answer
> we'd get is "yes, it might conflict and if you want to avoid any issues
> you wouldn't include it."

Exactly :), and I just had a discussion with some legal folks about
that because it has been an issue raised internally. So the boring
stuff being... The Perl License has two meanings: GPL v2.0 or Artistic
License 1.0, and there can be problems if fuzzystrmatch.so links with
something that has portions licensed as Apache v2 because Apache v2
and GPL v2.0 are incompatible, or anything who-know-what incompatible
with Apache v2. By choosing the Artistic license there are no problems
visibly, at least for the case I have been pinged about.

> To that end, I'd suggest -core simply formally ask the authors about it.
> Once we have that answer, we can figure out what to do.  In my
> experience, at least, it's a lot better to go that route and figure out
> what the original authors really *intended* than to go get a lawyer to
> weigh in on it.  One of those approaches is both free and gives a clear
> answer, while the other is expensive and doesn't provide any real
> certainty. :)

I would go this way as well, aka ask the authors and see if it is
possible to remove the license notice and keep everything licensed
under PostgreSQL license to avoid any future problems...
-- 
Michael



pgsql-hackers by date:

Previous
From: Andres Freund
Date:
Subject: Re: GSoC idea - Simulated annealing to search for query plans
Next
From: Amit Langote
Date:
Subject: Re: Partitioning WIP patch