Bruce Momjian <pgman@candle.pha.pa.us> writes:
> I think there was an updated BSD license approved by Berkeley that
> we are using.
I think this is an area where we need a higher degree of certainty
than that.
> If we took the file unchanged, I would not remove the copyright
> because it is the file _unchanged_, no?
It's the license, not the copyright attribution, that is the
question. I also don't see how it makes any difference whether the
files are modified (although at least some of these files _have_ been
modified since being imported: e.g. src/port/qsort.c).
To summarize, my understanding is that there are two problems:
(1) Some of the files in the main source tree are 4 clause
BSD. Since PostgreSQL is "derived" from these files, we fall
under its licensing restrictions, namely the advertising
clause.
We can solve this by getting the license on these files changed
to 3 clause BSD, or by removing the files from the
tree. Checking the current NetBSD CVS tree, it seems that they
switched to 3 clause BSD at some point after we imported the
files in question:
http://cvsweb.netbsd.org/bsdweb.cgi/src/lib/libc/stdlib/qsort.c
So it seems kosher to change the license on the NetBSD-derived
files to 3 clause BSD. That leaves the contrib/ stuff, but
since we still have GPL licensed code in there, I don't think
we need worry about it.
(2) Is the 3 clause BSD license identical to the PostgreSQL
license? It looks superficially dissimilar to me, but Bruce
thinks that ours is an "updated" version of the BSD license. If
that's the case we should be okay; if it's not, we may not
be. This needs to be confirmed.
(And, of course, IANAL...)
-Neil