[Pipet Devel] license issue revisited
J.W. Bizzaro
jeff at bioinformatics.org
Sun Sep 24 23:15:41 EDT 2000
Jean-Marc Valin wrote:
>
> I was actually discussing that with Richard Stallman yesterday... Here's what he
> had to say:
>
> "In a technical sense, the user does perform the act of combining them.
> But in truth they are inherently connected regardless of what any
> specific user does. They are designed to work as one program, and
> this is true before any specific user obtains his copy and runs it.
There is a problem, as he says, if Piper requires these libraries to run. I
don't believe that any of the libraries that Jean-Marc is concerned about will
be required. But, they should not be shipped/distributed with Piper either.
Then, it is the user's concern.
> I agree with him: we should play safe by explicitly specifying the exception,
> but the fact that the user does the linking, makes (I think) linking a non-free
> node with a GPL node legal. Since the linking is only done on demand at
> run-time, there's no redistribution problem.
Yes, I agree. We should include that exception. It doesn't mean, however,
that GPL and non-GPL libraries can legally be linked and then distributed
together. I think that we should also provide such a warning to users in the
license.
> Richard Stallman:
> Yes, that makes sense. You could say, for instance, that CORBA
> communication through the interfaces you have designed, is in your
> view interaction between two separate programs and does not make the
> communicating modules into a single program.
Yes, it makes sense to provide that exception, just for clarity's sake,
although it doesn't seem the GPL will make CORBA linking illegal anytime soon.
> Do we all agree?
I agree. Now, the question is GPL or LGPL ;-) I vote for GPL with the afore
mentioned exceptions.
Jeff
--
J.W. Bizzaro jeff at bioinformatics.org
Director, Bioinformatics.org: The Open Lab http://bioinformatics.org/~jeff
"Injustice anywhere is a threat to justice everywhere."
-- Martin Luther King, Jr.
--
More information about the Pipet-Devel
mailing list