Date: Thu, 03 Jun 1999 22:49:04 -0700 From: Peter Wemm <peter@netplex.com.au> To: Brian Somers <brian@Awfulhak.org> Cc: dyson@iquest.net, ahasty@mindspring.com (Amancio Hasty), dillon@apollo.backplane.com (Matthew Dillon), crossd@cs.rpi.edu (David E. Cross), freebsd-hackers@FreeBSD.ORG, schimken@cs.rpi.edu Subject: Re: 3.2-stable, panic #12 Message-ID: <199906040549.WAA01104@peterw.yahoo.com> In-Reply-To: Your message of "Fri, 04 Jun 1999 02:45:56 BST." <199906040145.CAA04373@keep.lan.Awfulhak.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Brian Somers wrote: > > It wasn't the "dark side" of core, it was the panic'ed and worried > > part of core that was seeing things happening without careful review. > > The system was becoming unstable due to Matts changes. Whether the > instabilities were in Matts code or somewhere else is irrelevent. > The reaction was (IMHO) the right thing to do. Well, having been a party to the discussions, I think I can pretty safely say that it wasn't so much a decision based on technical reasons but one based on an attempt at damage control. Tempers were boiling from all sides and things were getting rather ugly and going downhill fast. There was no good solution to the problem, and what was done in the end was (and I still think so) one of the least lousy of a limited set of choices that were even worse. Things had degenerated to the point that no matter what the outcome, it was a no win situation. I think the underlying problems were pretty fundamental and I suspect are still potential problems. I am going to talk about some of them. Keep in mind that these are my opinions, and I hope I'm not being unfair or biased given that I'm talking specficically about Matt's case. First, Matt had a lot of time to spend and the inclination and enthusiasm to do so. While this in itself is not a bad thing, it was a key source of friction between Matt and the other developers. Matt was keen to get things done *right now* and I suspect he found the delays in email communication rather intolerable, but that's the way the project (so far) has managed to keep together - with patience and consensis. A good number of developers *perceived* Matt to be ``running riot'' due to the pace he was going at (and a number complained about it), and toes were getting stomped on all over the place. Second, I got the impression Matt was more used to working alone or with a small, closely bound, group rather than a large distributed global project. Note that neither of these two are necessarily anyone's ``fault'', but more of an incompatability that wasn't being handled particularly well at all by any of the people involved. I still suspect the problems are still there and would be waiting to bite all over again if/when Matt gets commit privs back. I know Matt finds it frustrating working offline and via patches, but keep in mind that most of the developers already do this for larger changes. It's not uncommon at all for diff/patch sets to be posted for testing or work-in-progress snapshots. Unfortunately it's pretty much a necessary evil simply because of the geographic dispersion of people, timetables, timezones, time availability, etc. Some sort of arrangement/understanding/procedure/whatever would need to be worked out to make sure that everybody involved understands everybody's angle so that we don't repeat it all over again. Maybe some of the groundwork can be done at usenix next week, but not all everybody will be there. Anyway, I guess we can see. -Peter To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199906040549.WAA01104>