Legacy approval request - The PostgreSQL Licence

[Wilson, Andrew]

On Nov 30, 2009, at 10:36 AM, John Cowan wrote:
> Richard Fontana scripsit:
>> Are you confirming then that the license of PostgreSQL doesn't permit
>> sublicensing, even (somehow) by implication? I might conclude from that
>> that the license is GPL-incompatible, which is an issue that Tom tracks
>> for licenses that are approved for distribution by Fedora. (Not that
>> that has anything to do with OSI approval.)
>
> As far as I can see, the right to sublicense is irrelevant.
> An MIT/BSD-style license grants rights to anyone in lawful possession of
> the source code, however that possession came to be.  If Alice writes code
> under this license and sends it to Bob, then Bob can create derivative
> works and/or incorporate it into other software.  He has no need to
> reissue the code under his own license.

... unless he has combined the code with GPL code, which I believe
is Richard's point, since under the FSF interpretation of GPL
the entire combined work must be licensed under GPL.  In that
scenario an explicit right to sublicense, as in the canonical MIT/X11 license,
would be helpful.

I personally read this genre of BSD-like licenses as "everything which is
not forbidden is permitted."  Under that reading, combination of PostgreSQL
code with GPL code is allowed as long as the original copyrights and disclaimer of
warranty are preserved.

IANAL, TINLA, ho ho ho, merry Xmas.

Andy Wilson
Intel open source technology center