OVPL and open ownership

David Barrett dbarrett at quinthar.com
Mon Jul 25 08:16:50 UTC 2005


Chris Zumbrunn wrote:
> 
> Because while you see 3.3 as a "convenience" the stewards of the OVPL 
> see it as a "requirement". They are not willing to release the source 
> code without 3.3 being binding.

Perhaps I wasn't clear with my use of "opt out".  You can't opt out of 
following the license for existing code.  Rather, you have the option of 
not including 3.3 when licensing *new* code.

I'm not changing the fact that section 3.3 must be obeyed on all code 
licensed under it.  This includes at least all original ID code and all 
subsequent ID additions, but also all submissions by non-disgruntled 
contributors.  In other words, except in extreme situations (ID losing 
the faith of the community), virtually all code would continue to be 
licensed with 3.3 intact, from any author.  To repeat, under my 
proposal, the ID's exclusive privilege of relicensing code remains 
intact for everything he writes, and most of what he doesn't.  I can't 
overstate the importance of this point so please forgive the redundancy.

However, while contributors can't delete section 3.3 from the license 
covering existing code, they can choose to submit *new code* under a 
license that doesn't include it.  In other words, just as we're 
considering mandating submissions be made in BSD, I propose mandating 
that submissions be made in either the original OVPL, or OVPL without 
section 3.3.  Note again that just as contributing code under BSD only 
affects the licensing of the submission (it doesn't excise OVPL from 
existing code), contributing code under a choice of OVPL/OVPL' doesn't 
undermine the OVPL license of existing codes.  At worst, a contributor 
can choose not to grant 3.3 to the ID for any code he submits; he can't 
"undo" 3.3 grants from other contributors.

And I want to confirm that 3.3 is not a convenience; it's the sole value 
of the OVPL.  My whole intent is to ensure that *more* code is covered 
by 3.3 by making it the default submission license.  Indeed, my concern 
with the BSD proposal is that it *guarantees* all contributions will 
*not* covered by 3.3 -- a step in exactly the wrong direction.

Have I made this point clear?  If so (or if not), can you detail 
precisely why you believe my proposal does a worse job protecting the 
ID's privilege than licensing under BSD?  Can you give a concrete 
example under which it would be preferable to the ID to obtain 
contributions under a BSD license, versus under OVPL (with or without 3.3)?

(I think where I might have confused you is in combining this topic with 
an orthogonoal one about granularity of licenses.  I assumed files could 
only have one license, and thus I made conclusions based on this.  If we 
allow for multi-license files, then those conclusions don't count.  But 
given how license granularity equally affects contributions whether 
under BSD or OVPL', I'd like to separate that topic so as to avoid 
future confusion from that.)

-david



More information about the License-discuss mailing list