Re: GPLv3

[Arno Onken]

Thomas Weber wrote:
> Am Freitag, den 12.10.2007, 11:36 +0200 schrieb David Bateman:
>> Thomas Weber wrote:
>>> GPLv2 and later should be fine, shouldn't it? Or are there GPL2-only
>>> packages? 
>>>
>>> For other licenses, I don't know. But I'm not aware of any license that
>>> was compatible with GPL2 and isn't with GPL3.
>>>   
>> The issue is no that we can't transition to a GPLv3 license. 
> 
>> Rather the
>> issue that worries me is that the GPLv3 license states
>>
>> <quote>
>> the Free Software Foundation; either version 3 of the License, or (at
>> your option) any later version.
>> </quote>
>>
>> So officially at the moment you use a GPLv2 toolbox with Octave 2.9.15
>> you must effectively relicense it to be GPLv3. Can we assume this
>> process is automatic, or should we explicitly change the license of all
>> of the octave-forge code to avoid doubt...
> 
> If the toolbox is GPL2 only, you have a problem using it with a GPL3 (or
> later) Octave. If it says "GPL2 or later", the relicensing process is
> indeed automatic. 
> 
> Note that explicitely changing the license is practically not revertible
> (you'd need any later contribution to be licensed again back for GPL2). 
> 
> I don't know if it's worth the effort, unless there's something in GPL3
> which you need/like better than GPL2. For compatibility, "GPL2 or later"
> is sufficient. 

However, when scripts will be moved from octave to octave-forge in the
process of establishing the MATLAB-toolbox character in 3.1, many
octave-forge packages will inevitably be GPLv3 anyway.

Arno