From owner-freebsd-current Sat Apr 6 08:40:55 1996 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id IAA06946 for current-outgoing; Sat, 6 Apr 1996 08:40:55 -0800 (PST) Received: from rocky.sri.MT.net (rocky.sri.MT.net [204.182.243.10]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id IAA06931 for ; Sat, 6 Apr 1996 08:40:52 -0800 (PST) Received: (from nate@localhost) by rocky.sri.MT.net (8.6.12/8.6.12) id JAA21818; Sat, 6 Apr 1996 09:40:43 -0700 Date: Sat, 6 Apr 1996 09:40:43 -0700 From: Nate Williams Message-Id: <199604061640.JAA21818@rocky.sri.MT.net> To: paul@netcraft.co.uk Cc: freebsd-current@freefall.freebsd.org Subject: Re: sup/cvs tags In-Reply-To: <199604061346.OAA00485@originat.demon.co.uk> References: <199604052259.OAA19443@freefall.freebsd.org> <199604061346.OAA00485@originat.demon.co.uk> Sender: owner-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Paul Richards writes: > In reply to Mike Pritchard who said > > > > In the future can we try and avoid tagging the kernel source tree > > except at release time? The wollman_polling tag is causing anyone > > who sups the cvs files to receive every file in /usr/src/sys just > > because of the new tag. This looks to be about 33 megabytes of data, > > which is quite painful over a slow link. > > I agree. The FreeBSD cvs tree isn't for personal development and I'm > a bit miffed that Garret has been allowed to get away with putting in > a personal tag. I disagree. I think we don't use the full capabilities of CVS, and just because it might be painful for folks on slow links (and *NO* one has a slower link than me, so don't even try to argue that point), the benefits of being able to see the new code and possibility of getting newer code in the tree far outweighs the costs of increased transmission times. If it means we all switch to CTM (ugh, no offense Poul) then so be it. I'd *LOVE* to be able to apply the PC-CARD patches to a branch so that folks could get at them, and I could merge them into -current and -stable *AND* still allow folks to use them. As it stands now, they are dependant on me making a new (external) patch-set everytime I make significant changes to the code, which is a lot of work for me and a lot of work for them. With a branch on the tree much of this work would go away. Nate