From owner-cvs-src@FreeBSD.ORG Mon Aug 27 19:58:54 2007 Return-Path: Delivered-To: cvs-src@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id E5B2E16A417; Mon, 27 Aug 2007 19:58:54 +0000 (UTC) (envelope-from bright@elvis.mu.org) Received: from elvis.mu.org (elvis.mu.org [192.203.228.196]) by mx1.freebsd.org (Postfix) with ESMTP id CB2FC13C442; Mon, 27 Aug 2007 19:58:54 +0000 (UTC) (envelope-from bright@elvis.mu.org) Received: by elvis.mu.org (Postfix, from userid 1192) id 4E1251A4D97; Mon, 27 Aug 2007 12:56:34 -0700 (PDT) Date: Mon, 27 Aug 2007 12:56:34 -0700 From: Alfred Perlstein To: John Baldwin Message-ID: <20070827195634.GZ87451@elvis.mu.org> References: <20070824215515.GF16131@turion.vk2pj.dyndns.org> <200708270850.20904.jhb@freebsd.org> <20070827190100.GY87451@elvis.mu.org> <200708271529.42264.jhb@freebsd.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <200708271529.42264.jhb@freebsd.org> User-Agent: Mutt/1.4.2.3i Cc: src-committers@freebsd.org, peterjeremy@optushome.com.au, cvs-all@freebsd.org, deischen@freebsd.org, cvs-src@freebsd.org, Warner Losh , yar@comp.chem.msu.su Subject: Re: cvs commit: src/lib/libc/gen fts-compat.c fts-compat.h X-BeenThere: cvs-src@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: CVS commit messages for the src tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 27 Aug 2007 19:58:55 -0000 * John Baldwin [070827 12:37] wrote: > On Monday 27 August 2007 03:01:00 pm Alfred Perlstein wrote: > > * John Baldwin [070827 05:48] wrote: > > > > > > I think it will be confusing to have missing symbols just as folks would > > > have thought it confusing to have 6.x ship with libc.so.8 if we had > > > bumped libc multiple times. I also think that just managing the > > > interfaces that show up in releases and -stable branches will be enough > > > extra bookkeeping to keep track of as it is. > > > > This is something I just don't understand, why is some psuedo-arbitrary > > number somehow MORE confusing/damaging than some convoluted upgrade > > path? > > > > The only negative (which is bs) of doing so is keeping around multiple > > compat libraries for the same release, which we can simply decide not > > to do. > > > > (meaning, for your example there does not need to be a libc.so.7 > > shipped in some compat package unless someone really wants to) > > User questions. "I just upgraded via make world from 6.x to 7.0 and now I > have a libc.so.6 and a libc.so.8, what happened to libc.so.7, did something > go wrong??? please help!!!!" In the grand scheme of things, telling a user "that's expected" is a lot less overhead than actually performing triage on a "real problem". > etc. > > And yes, I do think it's ok for -current to have rougher edges. After all, we > aren't really trying to get people running -current on production systems. I don't know if we should be in effect punishing "power users", sure they can be annoying, but they really, really, really, really... do help with preventing regressions. Seriously, if only the extremely clued and brave can run current we are severely limiting our base of users who can test. I think this is more of a disservice to the project than the occasional power user whom we might actually have to strum up the *cough*balls*cough* to tell, "you know, maybe -current isn't right for you at the moment." -Alfred -- - Alfred Perlstein