Mozilla Public License 2 Alpha 3; request for early review prior to formal submission for approval

[Wilson, Andrew]

Rick Moen wrote:

> Quoting Andrew Wilson (andrew.wilson at intel.com):
>
>> Sorry, but in this scenario I clearly posit the case where Bob creates
>> not simply an archive but a linked executable which is distributed
>> under GPL, thus triggering GPL requirements to release the complete
>> corresponding source under GPL.  (to Carol)
>
> I'm with John Cowan:  This seems like just the urban legend that refuses
> to die.
>
> Specific performance is not even available as a remedy for copyright
> violation.  Plaintiff gets to enjoin further infringement, plus (if the
> work is registered) some degree of monetary damages.

Urban legend, my foot.

Expansive copyleft is what the black-letter text of GPL (both versions) actually says.
Expansive copyleft is what the license stewards of GPL, in their voluminous writings
and speeches, say they mean to achieve.  Is there any doubt of either point?

GPLv2 is approaching its 20th anniversary.  There has been no litigation, in
any jurisdiction, in 20 years, that has successfully challenged expansive copyleft.
Is there any doubt on this point?

Rick and John have their theories. Everyone has theories.
{I certainly do.}  However, unless and until the license stewards
change course, or courts force them to change, a realist (as opposed to a theorist)
will treat GPL as if it means exactly what it says,
and exactly what its authors say it means: expansive copyleft.

I will get off my soapbox now.

Meanwhile, back at least closer on topic for license-review: section 11.3 in the MPLv2
draft is specifically intended to address MPL code combined with GPL code.
It is apparently intended to augment, or perhaps supplant, the Mozilla tri-license.
It does this through the innovative mechanism of a license weak copyleft stand-alone
but permissive when combined with other types of licensed code.
So, it does seem reasonable to ask just how permissive combined-MPL is meant to be.

As a note, changing one word in section 11.3 from "distribute" to "sublicense" makes
combined-MPL as permissive as MIT or Apache.

Andy Wilson
Intel open source technology center