Re: Is there permissve GPL3 compatible license ?
W dniu śro, 19.08.2020 o godzinie 08∶05 +0200, użytkownik
tomas@tuxteam.de napisał:
> On Tue, Aug 18, 2020 at 09:17:36PM +0200, Marek Mosiewicz wrote:
> > Hello,
>
> hello
>
> > I'm just curious. According to general perception BSD, MIT or
> > Apache2
> > licenses are GPL3 compatible.
>
> There are many fine resources out there for you to read. I'd
> recommend [1], [2] and [3] for starters. Yes, [1] is gnu.org:
> one might think they have a bias -- still it's well structured
> and worth a read.
>
> > Is it true ? I'm not lawyer, but with GPL requirement to cover
> > whole
> > combined work with GPL it seems to be not so obvious.
>
> The thing is: BSD, MIT and friends allow you to combine the stuff
> with any other program and relicense the result as you wish (e.g.
> proprietary). This includes a GPL work. The result is, then, GPL.
BSD and MIT allows you to combine it with other work I agree. In MIT
license there is even permission to SUBlicense. Unfortunetly we are
talking about GPL (GPL3 especially)license and as from my understanding
it requires to publish WHOLE combined work on corresponding GPL
license. MIT/BSD license does not have any clue about changing license
in that way. That is fixed requirements and it is probably reason why
FSF even says GPL3 is not GPL2 comaptible. So how MIT/BSD could be
compatible.
>
> > That especially true for GPL3 where combined work must to be
> > covered
> > with GPL3, what means has requirements for patents used by work.
>
> That only means you lose the protection afforded by GPL3 once you
> assert patent rights (e.g. by submarine patents). Remember -- GPL
> /grants/ you permissions to do things which, by default, would be
> forbidden. So it only can revoke those things, if you are in breach
> of contract.
>
> > Neither BSD nor MIT licenses have any clue about author patents
> > used in
> > work. It seems obvious that you have patent grant from author as
> > long
> > as you use or modify BSD/MIT software,
>
> Is it? It will strongly depend on jurisdiction, I guess.
OK. I'm not sure of it. It seems for me that if you can use and modify
some software, that means that author gives You permission to use
patents in given work as long as they are authors patents (THAT IS ONLY
MY ASSUMPTION). That of course does not cover any patents not belonging
to software authors
> > but nothing about any other use
> > of patents.
>
> Why should the patent grant be tied to the license of the end
> software? Now, if the license explicitly states so, perhaps.
>
> > I'm not sure for Apache2. There is patent grant from contributors,
> > but
> > I do not see patent grant from copyright holder (who can be not
> > contributor, but also have patents)
>
> The patent grant in Apache2 is why the FSF recommends this one
> among the non-copyleft [4] licenses.
>
Is it GPL3 compatible ?
>
> [1] https://www.gnu.org/licenses/license-list.en.html
> [2]
> https://en.wikipedia.org/wiki/License_compatibility#Compatibility_of_FOSS_licenses
> [3]
> https://en.wikipedia.org/wiki/Comparison_of_free_and_open-source_software_licences
> [4] I very much prefer this term to the "permissive" you used in the
> subject: permissive is ambiguous. Permissive to whom? Users?
> Programmers? Distributors?
>
> - t
Reply to: