|
Boost : |
Subject: Re: [boost] BOOST license & GPL
From: Niklas Angare (li51ckf02_at_[hidden])
Date: 2017-04-01 22:13:43
"Andrey Semashev via Boost" wrote:
> On Sat, Apr 1, 2017 at 9:22 PM, Niklas Angare via Boost
> <boost_at_[hidden]> wrote:
>>
>> But if they distribute the whole thing saying "this is licensed under the
>> GPL", doesn't that kind of contradict the requirement to include the
>> Boost
>> Software License and the copyright notices?
>
> It does, if their distributed source (as a package or a VCS snapshot)
> does not contain BSL with a clear indication of what code is under
> BSL.
I can't find the Boost license in the KiCad source.
>> It would feel better to me if
>> they said "this is licensed under the GPL and, in part, other compatible
>> licenses".
>
> It doesn't work like that. A product, as a whole, must be distributed
> under a single license. (It can be distributed under multiple
> licenses, e.g. proprietary and copyleft, but that is another off-topic
> case; each such distribution is independent and whole.)
Why does the whole source package have to be distributed under a single
license?
> As I understand it, a license, among other things, defines two sets: a
> set of rights granted to the user, and a set of restrictions or
> obligations imposed on the user. Two licenses are compatible if the
> sets defined by the licenses are non-contradictory and can be
> fulfilled combined. BSL is less restrictive than GPL, and in order for
> the two of the licenses to be fulfilled the code must be distributed
> under the GPL. It doesn't make that BSL part on its own licensed under
> the GPL (or rather, you cannot enforce GPL on that part because the
> original copyright holders did not allow that).
The requirement that the entire BSL be included is not a part of the GPL, so
for that particular detail the BSL is more restrictive than the GPL. Maybe
I'm nitpicking.
Regards,
Niklas Angare
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk