From owner-freebsd-current Sun May 9 15: 6:17 1999 Delivered-To: freebsd-current@freebsd.org Received: from nomad.dataplex.net (nomad.dataplex.net [216.140.184.132]) by hub.freebsd.org (Postfix) with ESMTP id 4E99C14D95 for ; Sun, 9 May 1999 15:06:13 -0700 (PDT) (envelope-from rkw@dataplex.net) Received: from localhost (rkw@localhost) by nomad.dataplex.net (8.9.2/8.9.3) with ESMTP id RAA83100; Sun, 9 May 1999 17:02:57 -0500 (CDT) (envelope-from rkw@dataplex.net) X-Authentication-Warning: nomad.dataplex.net: rkw owned process doing -bs Date: Sun, 9 May 1999 17:02:57 -0500 (CDT) From: Richard Wackerbarth Reply-To: rkw@dataplex.net To: Poul-Henning Kamp Cc: Chuck Robey , freebsd-current@FreeBSD.ORG Subject: Re: kernel.old In-Reply-To: <21918.926281586@critter.freebsd.dk> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG I wish :-( It seems that some people think that it is OK to make changes to stable even though those changes break things which used to work. IMHO, branches of the kernel SHOULD be like shared libraries. (It is OK to ADD previously absent features or CORRECT internal errors, but NOT OK to delete features or change API's) On Sun, 9 May 1999, Poul-Henning Kamp wrote: > I think you are seeing -current as the norm. You shouldn't. Under > -stable the modules should (tm) continue to work since there are not > made API changes in -stable. Personally, I think that we should treat kernels just like another library. They export an API (sysctl) that libc, et. al. uses and another API that the kernel modules use. Any change that "breaks" code which is compliant with those API's belongs in a new release branch. PERIOD. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message