Date: Tue, 3 Sep 1996 12:19:23 -0700 (MST) From: Terry Lambert <terry@lambert.org> To: phk@critter.tfs.com (Poul-Henning Kamp) Cc: msmith@atrad.adelaide.edu.au, rkw@dataplex.net, jkh@time.cdrom.com, current@FreeBSD.org Subject: Re: Food for thought Message-ID: <199609031919.MAA04862@phaeton.artisoft.com> In-Reply-To: <6310.841765611@critter.tfs.com> from "Poul-Henning Kamp" at Sep 3, 96 05:46:51 pm
next in thread | previous in thread | raw e-mail | index | archive | help
> Listen, it's really VERY VERY SIMPLE! We don't promise anything about > -current, nothing, absolutely NOTHING! > > If you have trouble with -current, you'd better find out for your self > what they are, and send an email with a patch, or at least >very< precise > description to the committer you think is responsible. > > Otherwise: Get of our back! The problem is not one of meeting promises which you have outstanding; I will be the first to agree that there are no promises out there which you should feel compelled to meet, in regard to -current. On the other hand, I think that there is general agreement the overall group effort can be made to move more effectively toward the combined project goals if -current compiles. I'm not suggesting that it would then have to *run*. *That* would be trying to make -current into "production level code", which I agree, would be a mistake. On the other hand, every 3-4 months, -current goes through a period of instability, we all have this very argument, people get their sensibilities offended, and then they set about proving we don't need any "grandios schemes" by increasing their own dilligence. And things improve for 3-4 months until. once again, the dilligence slips. The problem is that there is a human dilligence requirement that needs to be removed from human control and ensconced in the process itself so that a human failure will not cause a failure of credibility. I have made my own suggestions on this several 3-4 month periods in the past, on several occasions, already. I'd be perfectly happy to see *any* soloution: it does not have to be mine, and I'll stop advocating mine if it will help pave the road to *some* soloution. Frankly, I find it apalling to have the *possibility* of a source repository *ever* in an unbuildable state following a group of related checkins. It simply irks me no end, since it violates a lot of my precepts of professionalism. Yes, my suggested fix would put a build latency into the process of checkin. I am perfectly happy to hear other proposals for fixes that wouldn't. Meanwhile, if nothing else, we have bought 3-4 more months in which to address the problem or prepare for another session to get ourselves another 3-4 months after that... Regards, Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199609031919.MAA04862>