On Tue, 9 Mar 1999, Martin Husemann wrote:

> Greg A. Woods wrote:
> 
> So the only way to keep it simple is to require *without exceptions* all
> stuff to be assigned to TNF.

I hope that this does not become the case. A number of NetBSD developers [5] 
work for me at NASA Ames and regularly contribute code written while in the
pay of NASA. Since this issues recently came to the forefront of discussion,
we will have to evaluate if we can contribute any more changes in the future.

I can't speak for NASA officially, but can easily see this going in a
direction that may be very bad for all of us. For one thing, we'd ideally like
to be able to allow use of what we develop to whoever we want. e.g. FreeBSD
or [pick your favorite version of] Linux. Unfortunately, NetBSD does not
have a user base as large as some other OS's. When we attempt to "cooperate"
with some random vendor, we usually tell them that we plan to make the details
of the drivers available to other OS's, and this seems to make things more
attractive to them when we try to get internal docs, etc... Oh yea, it's also
really hard to just give things away from nasa. [small things, like a bug fix
can be just given away, but if larger pieces of work can not just be _given_ 
to NetBSD without a lot of extra paperwork]

So, hopefully, some exception mechanism can be found that is agreeable to
everyone. I just thought I'd raise this as an example of what types of
problems may/will come up. 

Harry Waddell
waddell@nas.nasa.gov

p.s. Please don't confuse the above with an official NASA position. It's
actually just mine.