Scripsit Raul Miller <[EMAIL PROTECTED]>
> On Thu, Jan 13, 2005 at 09:18:21PM +0000, Henning Makholm wrote:
>> To make what I fear explicit, here is a fleshed-out scenario:
>> 1. A writes a program and releases it under the current CPL.
>> 2. B takes A's program, hacks on it, distributes his Contributions
>> on a website under the current CPL.
>> 3. IBM notices B's modified program and decides that they would like
>> to include it in their proprietary program suite for frobitzing
>> foobars, with private modifications that they don't release
>> source for. So,
>> 4. IBM releases new version of CPL which gives IBM carte blanche to
>> do anything at all with covered code.
>> 5. C (a strawman for IBM) picks up B's modified program, makes some
>> inconsequential little changes and releases it under the new
>> amended CPL - as allowed by the original CPL under which B
>> distributed his changes.
>> 6. IBM picks up C's distribution and starts abusing B's contribution
>> proprietarily.
> The same thing is possible with the GPL, with it's "any later version"
> clause.
No - the GPL would allow B to distribute his hacked code under "GPL,
version 2 only" if he wants to prevent this.
> For that matter, the same thing is possible with the BSD license,
> because it makes no attempt to provide copyleft protection.
No - the BSD license would allow B to distribute his hacked code under
"GPL, version 2 only" if he wants to prevent this.
The CPL does *not* allow B to distrubte his hacked code under "CPL,
version 1 only". In fact, "CPL, version 1 only" is not even a
internally consistent license specification, because CPL
*intrinsically* allows any downstream recipient to switch everything
to any later version published by IBM.
--
Henning Makholm "Jeg k�ber intet af Sulla, og selv om uordenen griber
planm�ssigt om sig, s� er vi endnu ikke n�et dertil hvor
ordentlige mennesker kan tillade sig at stj�le slaver fra
hinanden. S� er det ligegyldigt, hvor st�rke, politiske modstandere vi er."
