[License-review] CC0 incompliant with OSD on patents, [was: MXM compared to CC0 ]

[Tzeng, Nigel H.]

On Thu, 8 Mar 2012 12:56:30, Bruce Perens <bruce at perens.com> wrote:
  
>>On 03/08/2012 07:32 AM, Tzeng, Nigel H. wrote:
    
>>As long as it's plainly stated that the license does not provide a
>>patent
>>license then users of the software can judge for themselves the
>>risks
>>involved. Just like they do with copyleft vs permissive licenses.

>They can judge the risk regarding copyright-based licenses because
>the copyrighted material is there in front of them. This is not the
>case regarding patents.
 
It doesn't matter.  My assertion is that simply folks that judge any
patent risk as too high will simply avoid the use of any software
that does not provide an explicit patent grant.  They can do this
without any knowledge other than what the license says.

This is exactly the same risk analysis regarding any uncertainty with
copyleft.  Folks that do not wish to worry about the linking controversy
or whatever will simply ignore GPL licensed software and use only LGPL,
Apache, BSD, etc. 
   
>If I were to evaluate the U.S. patent risk for a company, a great
>deal of the burden would be search and evaluation of a very large
>collection of patents. The engagement would be around USD$100K, and
>that's not counting what the attorney who gets my report would
>charge. And there would be some big caveats in my report.
    
>The cost of engagement for evaluating risk and establishing policies
>and processes for combining Open Source and proprietary software
>would be less than 1/5 of that.
 
The costs and results are very similar.

These licenses are approved for use:  <list of approved licenses>
    
 
>>Frankly, it will likely be corporations that will avoid the use of
>>open source without explicit patent grants more often as a matter
>>of policy.

>No, they will prefer to use software with such grants
>because it reduces their risk.

That is exactly what I said albeit with a confusing double negative.

>What they worry about is contributing their own work to such software.
>In general the ones who also have a patent royalty business either
>refuse to make such contributions or use intermediaries to develop them.
   
Yes.  Or they use some other license that provides source with only
limited or no patent licenses.  Since there are no commonly used
licenses that provide these in the "open source" world they will likely
be different, confusing and fragmented.

That may be awesome for open source purists.  Not so much for the rest
of the developer community.
 
>If you want an organization that recommends licenses, the FSF is
>happy to help. I agree that OSI should have a short-list of
>recommended licenses, but the politics of dis-recommending some
>organization's license are too much for them.

If I wanted a very narrow view of what open source is then yes, the
FSF is satisfactory.  I do not.

Frankly, what I want is the family of CC licenses for software.  Then
devs without some philosophical axe to grind will have a wide range of
very permissive to moderately restricted software common licenses
to use as needed on OUR terms.

Nigel