Creative Commons style open source licenses?

[Grant Robertson]

I have only been reading this list for a short time. However, it really seems to me that the OSI needs to sit down and do for open source licensing what Creative Commons has done for general copyright licensing. The people of the OSI need to sit down and list all the different features that people want in an open source license. All the different rights that an original developer wants to give to users of their code. All the different responsibilities that using such rights will require. Then the OSI needs to create a system where an original developer can easily just check off the rights and responsibilities they want to confer to users of their code and quickly create just the license they need. Each of these rights and responsibilities could be encoded so that the full text of the license wouldn't need to be included in each and every tiny piece of source-code that was used. Developers could just include a URL to the OSI web site with a code indicating the type of license. Perhaps even a snippet of XML allowing different responsibilities to be listed for different rights. So, in-house use could require a certain set of responsibilities, commercial use could require a different set of responsibilities, and open distribution could require yet a different set of responsibilities. In essence, an XML standard for licensing, if you will.

 

The system would know which of the different licensing terms were contradictory or incompatible and warn whoever was building the license of this, or preferably prevent those licenses from being generated at all. The system could also allow subsequent developers to add or remove rights or responsibilities as long as those changes were compatible with the original license. Rather than needing to pour over a license to see if it were legal to add or remove a right or responsibility and debate it endlessly in this forum, a subsequent developer could simply go to this web site and enter the snippet of XML given in the original source-code. The web site would then present them with a list of what rights and responsibilities they could modify to grant more or fewer privileges to any other developers subsequent to them. The web site could even allow the user of any open-source code to enter the license codes (XML snippets) of all the different pieces or open-source software they were thinking of using for their project and it would instantly tell them if the licenses for all those pieces of source-code would be compatible for use within the same derivative work.

 

It might even be possible for this "system" to be an XML schema designed such that anyone could use it to build or modify a license with basic XML editing software. Finally, an XSLT could be created that converts the license XML "documents" into full text licenses. This would provide consistency while preventing self-contradictory licenses.

 

I think this would be much better than the mish-mash of licenses that currently exist. These first generation licenses were created in a vacuum, so I can see why the first few of them were written as independent works with different and contradictory wording. But it has been 18 years since the first GNU license was released. I would think that a community of software developers would have come up with something more standardized than just saying "Apache + limitations" or "MIT + BSD" or whatever, by now.

[John Cowan]

Grant Robertson scripsit:
> I have only been reading this list for a short time. However, it
> really seems to me that the OSI needs to sit down and do for open
> source licensing what Creative Commons has done for general copyright
> licensing.

That worked for CC because it was a greenfield situation: there were
no open licenses for general content, and they could create a license
templating system from scratch (although drawing on the FLOSS licenses
to tell them what options made sense).  Such was not the case when the
OSI was founded.  Attempting to replace the existing dominant licenses
with new ones at this point would be quixotic.

--
Híggledy-pìggledy / XML programmers            John Cowan
Try to escape those / I-eighteen-N woes;        http://www.ccil.org/~cowan
Incontrovertibly / What we need more of is      cowan@...
Unicode weenies and / François Yergeaus.

[Arnoud Engelfriet-2]

John Cowan wrote:
> That worked for CC because it was a greenfield situation: there were
> no open licenses for general content, and they could create a license
> templating system from scratch (although drawing on the FLOSS licenses
> to tell them what options made sense).  Such was not the case when the
> OSI was founded.  Attempting to replace the existing dominant licenses
> with new ones at this point would be quixotic.

One thing OSI could do is present a categorization like CC has
created for open source licenses. Right now the list of "popular
and widely used licenses" is a mere list.

For example, subheadings with "no copyleft", "weak copyleft" and
"strong copyleft" would be easy to insert and would be very helpful.

Arnoud

--
Arnoud Engelfriet, Dutch & European patent attorney - Speaking only for myself
Patents, copyright and IPR explained for techies: http://www.iusmentis.com/

[Allison Randal]

John Cowan wrote:
  >
> That worked for CC because it was a greenfield situation: there were
> no open licenses for general content, and they could create a license
> templating system from scratch (although drawing on the FLOSS licenses
> to tell them what options made sense).  Such was not the case when the
> OSI was founded.  Attempting to replace the existing dominant licenses
> with new ones at this point would be quixotic.

True enough, but given a 30-year view rather than an immediate-future
view, it's a really good idea. It may ultimately be the best path out of
the compatibility mess. I've been saying this for a couple of years now,
and in the past year have seen it popping up independently on multiple
different blogs, which is a good sign that it's not only an idea that's
right for the time, but one with a good chance of adoption. I'm not sure
OSI is the right organization to do it, though.

Allison

[Ernest Prabhakar]

Hi Arnoud,

On Apr 22, 2007, at 3:44 AM, Arnoud Engelfriet wrote:
> One thing OSI could do is present a categorization like CC has
> created for open source licenses. Right now the list of "popular
> and widely used licenses" is a mere list.

I know this is something the OSI has long wanted to do. The challenge  
has been the implementation. Hopefully this year...

-enp