Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!hookup!swrinde!ihnp4.ucsd.edu!agate!agate.berkeley.edu!cgd
From: cgd@erewhon.CS.Berkeley.EDU (Chris G. Demetriou)
Newsgroups: comp.os.386bsd.misc
Subject: Re: Impressions: FreeBSD vs Linux
Date: 24 Mar 94 14:29:44
Organization: Kernel Hackers 'r' Us
Lines: 44
Message-ID: <CGD.94Mar24142944@erewhon.CS.Berkeley.EDU>
References: <1994Mar18.084355.19503@atlas.com>
<CMzw69.92K@tower.nullnet.fi><Cn1KJ1.9pr@boulder.parcplace.com>
<HJSTEIN.94Mar24111940@sunset.huji.ac.il> <xs-NJYB.dysonj@delphi.com>
NNTP-Posting-Host: erewhon.cs.berkeley.edu
In-reply-to: John Dyson's message of Thu, 24 Mar 94 16:26:25 -0500
In article <xs-NJYB.dysonj@delphi.com> John Dyson <dysonj@delphi.com> writes:
[ the way people 'take' 'assignments', etc. ]
I'd say that's a rather good description of it.
Basically, for NetBSD, what happens is the following:
The people that want to do work to be put in the master source tree
coordinate with the people 'in charge' of the given project, who
provide guidance, support, etc., as necessary. When they're done,
the changes go into the source tree, and usually are accompanied
y a short description in a file.
That is, for many things, there's little discussion in any forum with
the general public.
The person working on a project more or less 'owns' it, and if somebody
volunteers for something, we *expect* to get something from them...
For things which are 'in the works', but relatively far off, developers
are left to pursue them independently, and, generally are encouraged
to be quiet about what they're doing -- that way:
(1) they get less mail about it
(2) the people 'in charge' become a clearinghouse of
ideas and projects, and people know to talk to
*them* to find out who's working on something
(or whether they should do so, etc.)
It leads to a much less verbose, by much more tightly integrated
development gorup.
I'd much rather see more-or-less centrally-coordinated system,
like the ones NetBSD and FreeBSD currently have, than a relatively
ragtag effort; i think it's more efficient (in terms of amount of
good code produced), wastes less bandwidth, and generally produces
a 'tighter' system of software.
cgd
--
chris g. demetriou cgd@cs.berkeley.edu
you can eat anything once.