Re: IBM Public License (again)

[Raul Miller <moth@debian.org>]

> > I agree -- and maybe I'm stupid, but I don't see a contamination
> > mechanism here. [...]

On Fri, May 14, 2004 at 09:33:31AM +0100, MJ Ray wrote:
> It imposes restrictions on what actions you can take over other 
> software.

That might make it incompatible with the GPL, but this is a typical
characteristic of many licenses -- including DFSG licenses.

[Another example: a "must rename" clause is an additional restriction
on what actions you can take over other software.]

But the question is: how do these restrictions contaminate other software?
[Contaminate other software means that just being on the same media is
a problem.]

> > Can you point me at a contaminating mechanism which isn't also present
> > in a BSD license?
> 
> The patent paragraph of section 7?
> 
> With a BSD-style license, we would not have previously taken a license 
> for patents held by the holder. Isn't number of licensees in similar 
> situations taken as evidence in court?

But how is this contamination?

If there were any valid patent licenses covering this software, under
BSD you'd be in trouble for patent violation.

If there are any invalid patent licenses covering this software, these
clauses give you protection from litigation based on those licenses.

If there are no patent licenses, this is indistinguishable from BSD's
treatment of those licenses.

Or maybe I'm missing something?

Thanks,

-- 
Raul