Issues with rtl_tcp

Sylvain Munaut 246tnt at gmail.com
Thu Sep 27 00:24:30 UTC 2012


Hi,


> I think you misread Leif's message - what was said is correct.  You can take
> an LGPL library and relicense it under the GPL, but you can't go back to
> LGPL again.  Yes, this implies the license becomes more restrictive, because
> there are things you can do under the LGPL that you can no longer do once it
> becomes GPL.

I think you misread it. He clearly stated "the library license is more
restricted and
that things you are allowed to do under the library license are still
allowed under the general license."

This is not the case. In LGPL you can link so closed software, in GPL
you can't. So everything allowed under LGPL is _not_ allowed under
GPL.


> I think what Leif was getting at is that you can do whatever you want if you
> don't distribute it.  The GPL and other licenses only come into effect once
> you distribute something to someone else.

Well yes, obviously if you want to write source code just for you and
never distribute it (not even the source) that's totally fine
obviously.



> If I understand correctly, Leif was suggesting a workaround.  Instead of
> distributing something that violates the license, you could distribute
> something in two halves (each in compliance), along with instructions on how
> to link them together.  Once linked they can no longer be distributed, but
> because you were distributing them separately you wouldn't be violating any
> licenses.

Distributing code source that cannot work without the GPL library is
not compliant.
GPL doesn't mention linking explicitely, just derivative work and a
piece of source code that can't be compiled into an executable without
the GPL piece of code linked to it would definitely be considered
derivative work. If not, the whole concept of "gpl compatible"
open-source license wouldn't make sense.

But in any case, in his case it's completely irrelevant because the
open source license he chooses _is_ GPL compatible so there is no
issues to start with.


Cheers,

      Sylvain




More information about the osmocom-sdr mailing list