License Proliferation

Chris Zumbrunn chris at
Sun Sep 4 23:11:43 UTC 2005

On Sep 4, 2005, at 10:10 PM, Alex Bligh wrote:

>> I understand why you are saying that. But this is exactly what I think
>> would not
>> be the case. Trying to force the concepts of both "open" and "free"
>> licenses under
>> a single umbrella will always be trouble. Both are cool, but not the
>> same. Both
>> are companions, but unfortunately only one is compatible with the 
>> other.
> We must be at cross purposes. What are the definitions of "open" and 
> "free"
> you are using?
>> From you second sentence, I am guessing (but see below) that you are
> referring to reciprocal licenses as "open" here, and what Larry calls
> academic (or BSD-esque) licenses are "free". Whatever one's views on 
> these
> license types, I think that is not the common use of the terms, and 
> only
> going to cause confusion.
> I don't think you can use the word "open" in a manner that *excludes*
> BSD-esque licenses (academic licenses) - I think there's simply no
> precedent for that. (or indeed excluding reciprocal licenses).
> I don't really like the word think "free" is a great term for license
> taxonomy, as:
> 1. To many people it implies "does not cost money". That includes
>   binary-only freeware...
> 2. ...but to the FSF et al. it implies a political statement (no
>   problems here), which does not result in a set of OBJECTIVE criteria
>   (which means whilst it's useful, it doesn't help with license 
> taxonomy).
> It's thus confusing, and adds nothing, but in any case I fail to see 
> how
> you could describe either the GPL, or the BSD license, as non-free 
> (which
> you seem to imply above)
> I don't like the word "compatible" (which implies symmetry) either 
> though
> I can hardly criticize you for that "only under license type can
> code be relicensed under the terms of the other" is presumably the
> point you are trying to make.
>>> The GPL not being a recommended license would conceivably be a
>>> maintainable
>>> position, if carefully explained. (For instance, if the GPL 2.0 was 
>>> not
>>> a recommended license, but 3.0 was, that would clearly be an 
>>> acceptable
>>> position).
>> Yes. Hopefully, if we fast-forward a few years, "open" and "free" code
>> will be
>> bi-directionally compatible. Making the distinction now would help us 
>> get
>> there.
> ... but now I am totally confused. Assuming I guessed right what
> you mean by "free" and "open", how will code that continues to be 
> licensed
> only under a reciprocal license ever be relicensable under an academic
> license?

Sorry, I'm knowingly over-simplifying the issue and I'm assuming that
others read the same meaning into the terms "open" and "free" as I do.
The reality of course is not as black and white as I'm making it to be
right now.

With "free" I mean licenses that only allow sub-licensing of larger 
works under
the same license (GPL). Licenses that free the code from the threat of 

With "open" I mean licenses that allow sub-licensing of larger works 
even under
a proprietary license (BSD). Licenses that open the code for 
exploitation by others.

The reality is of course no where near as clear cut, but the 
simplification is
intentional and I'll stick with it, because my point is that it would 
be in
everybody's interest to make that distinction more prominently. The two
concepts are like oil and water - they do not mix.

The FSF has a very clear idea about what it considers "free" and what 
The overall open source community on the other hand generally tries to
define itself in ways that includes both "free" and "non-free" open 
licenses and as a result portrays an image to the general public that 
easy to understand at first glance but gets very confusing as soon as 
look closer.

My point is that if we define "open" licenses to be combinable, 
allowing the
creation of larger works under any other license, including a 
proprietary one,
then the two open source concepts would become easier to undertand for
the users and licensees and that would be good for the entire open 

"Open" and "free" would become the Yin and Yang of the open source
movement and the clear difference will actually make it easier to 
compatibility and harmony.

If "code reuse" is truly the main objective of the license 
proliferation concern
then the OSI could/should really make a bold step in that direction and
make license compatibility a condition in the OSD.


More information about the License-discuss mailing list