From owner-freebsd-stable Thu Oct 2 13:39:28 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id NAA21327 for stable-outgoing; Thu, 2 Oct 1997 13:39:28 -0700 (PDT) Received: from sabre.goldsword.com (sabre.goldsword.com [199.170.202.32]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id NAA21322 for ; Thu, 2 Oct 1997 13:39:25 -0700 (PDT) Received: (from jfarmer@localhost) by sabre.goldsword.com (8.8.7/8.7.3) id QAA03980; Thu, 2 Oct 1997 16:41:37 -0400 (EDT) Date: Thu, 2 Oct 1997 16:41:37 -0400 (EDT) From: "John T. Farmer" Message-Id: <199710022041.QAA03980@sabre.goldsword.com> To: jkh@time.cdrom.com, kkennawa@physics.adelaide.edu.au Subject: Re: CVSUP vs. SNAPS Cc: andrsn@andrsn.stanford.edu, dg@root.com, freebsd-stable@FreeBSD.ORG, jfarmer@goldsword.com, rgrimes@GndRsh.aac.dev.com Sender: owner-freebsd-stable@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Thu, 02 Oct 1997 10:24:56 -0700 "Jordan K. Hubbard" said: > [Others said...] >> I should also note that at that time the web documentation still referred >> to 2.1 as the stable branch, and 2.2 as current, which did not help to >> ease my confusion. I should HOPE this has changed by now :) > >It should be, though if you find any instances where it's still not, >please point them out to www@freebsd.org! :) > >> So if my opinion counts for anything in this debate, it goes for the >> 2.2.5-stable branch name. :) > >But the branch is called RELENG_2_2 - that is physically the name of >the tag. There is no RELENG_2_2_5 branch tag, which is what your >suggestion would imply and cause even more confusion. > >I think we're really better off just leaving it the heck alone for >now. During BETA test is *not* the time to contemplate major changes >in our release engineering strategy, for better or for worse. This >should have been brought up several months ago if Rod was actually >hoping for any sort of genuine effect here. :-) As it is, I certain >intend on doing absolutely nothing different than "usual" at this late >stage in the game. > Well, I suspect that Rod thought he had "a genuine effect" back at the 2.1 stage!! Of course , I remember the _same_ argument occuring at the 2.1.5 to 2.1.7 beta/release stage... I think the fact that the argument occurs at _every_ major release point should point out that the terms Jordan wants to use are at odds with a substantial number of users. Heck, having the labeling roll back like that is at odds with _everything_ I've been taught & have taught about life-cycle management in my 17+ year career... If it's the Beta release for the 2.2.5 version, then it should be labeled as the 2.2.5 Beta. The branch label & tags should reflect that. Effectively when you go into the beta cycle, ALL commits to the 2.2-STABLE tree should stop, UNLESS you can guarantee simultaneous commits to BOTH trees. Because, if you insist on continuing the old tree along with the new, they are seperate branches. Yes, yes, I know that they're the same code, but the labeling _MUST_ be different. Otherwise you can never be certain what you're dealing with. But it another way: will there _ever_ be another release based on 2.2.2? No. The date you declared the 2.2.5 beta test period to be in effect has frozen 2.2.2 in time, with only severe bug fixes to be applied. Any new releases from the 2.x strain _will_be_ based on 2.2.5. Remember, cvs, branch labels and tags are there to support release engineering & management, not to force you into a one-size fits none mold. John ------------------------------------------------------------------------- John T. Farmer Proprietor, GoldSword Systems jfarmer@goldsword.com Public Internet Access in East Tennessee dial-in (423)470-9953 for info, e-mail to info@goldsword.com Network Design, Internet Services & Servers, Consulting