Message-Id: <199512200400.FAA06746@atlas.fwi.uva.nl>
From: frank@fwi.uva.nl (Frank van der Linden)
Subject: Re: What can one check in?
To: netbsd-developers@netbsd.org
Date: Wed, 20 Dec 1995 05:00:32 +0100 (MET)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1251      
Sender: owner-netbsd-developers@NetBSD.ORG
Precedence: first-class

I've wondered about 'what to check in' myself a couple of times, since
I hope to contribute more than just Linux compat code.  It was never
serious, since whatever I did outside my little compat_linux domain was
quite easily handled by running it by the person responsible for it and then
checking in, or simply notifying the person responsible so that he could do
it himself (a preferred method, I think).

Generally, it's best if any ventures outside one's trusted domain would
go via core or another person responsible. But that might create a
bottleneck and slow down things a bit. The other end is to trust each
developer to do the Right Thing. I think that what would indeed be
important (as I've said before, and as Ted noted) that a list of
who's working on what is maintained somehow. That way, 'core' can
look at projects that dig deeper than just fixes, and steer them
in the right direction by staying in contact, which would make
trusting people to do the right thing easier, since 'core' would
have had enough influence on any bigger project by the time it's ready
to be checked in (substitute 'port-master' for 'core' on any architecture
specific project).

Main point: a "who's-doing-what"-list would be a good thing to have.

- Frank