From owner-freebsd-arch@FreeBSD.ORG Tue Mar 30 21:28:14 2004 Return-Path: Delivered-To: freebsd-arch@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 2386A16A4CE for ; Tue, 30 Mar 2004 21:28:14 -0800 (PST) Received: from gavriel.barak.net.il (gavriel.barak.net.il [212.150.48.102]) by mx1.FreeBSD.org (Postfix) with ESMTP id 4D32D43D1F for ; Tue, 30 Mar 2004 21:28:13 -0800 (PST) (envelope-from alex@hightemplar.com) Received: from hydralisk ([62.90.149.2]) by gavriel.barak.net.il e5bc39f1001e7dfa47fa92d56cd12779) with ESMTP id <20040331052730.YDLR4642.gavriel@hydralisk> for ; Wed, 31 Mar 2004 07:27:30 +0200 From: Alex Keahan To: freebsd-arch@freebsd.org Date: Wed, 31 Mar 2004 07:27:48 +0200 User-Agent: KMail/1.5.1 References: <20040328094048.GA40406@phantom.cris.net> <20040330232429.GA65170@phantom.cris.net> In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Message-Id: <200403310727.48043.alex@hightemplar.com> Subject: Re: CFD: XMLification of NOTES X-BeenThere: freebsd-arch@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussion related to FreeBSD architecture List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 31 Mar 2004 05:28:14 -0000 On Wednesday 31 Mar 2004 1:54 am, Dag-Erling Sm=F8rgrav wrote: > Alexey Zelkin writes: > > On Tue, Mar 30, 2004 at 05:05:24PM +0200, Dag-Erling Sm?rgrav wrote: > > > NOTES is not intended as a list of supported device and options. We > > > have src/sys/conf/files* and src/sys/conf/options* for that. > > > > Strange. IMO LINT was existed for exactly this reason > > Your opinion does not matter. The purpose of LINT is to cover as much > code as possible. Why do you think it's called LINT? Why doesn't his opinion matter? Perhaps my opinion does not matter either, but as a developer and long-time FreeBSD user, I don't think that XMLification of NOTES is such a bad idea. Having said that, I can also see why some developers would be opposed to the idea. Apart from neophobia and an obligation to fight to keep Unix pure, simple and free from modern-day buzzword nonsense, the main reason for the hostility and ridicule with which Alexey's idea was treated = is the simple fact that an XML kernel configuration tool would likely shift the blame for misconfigured kernels from the end-user to the developer. Whereas previously a developer could add a new option or remove an existing one without having to think twice about it, a strict XML schema would force the developer to adhere to higher standards, explicitly evaluate and list the dependencies and possible incompatibilities and reconcile the new options with the multitude of the existing configuration possibilities. A good thing for the end-user? Certainly. At the moment, a user who has the misfortune of breaking his system by adding an "option FOO" to his kernel without the required corresponding magical "option BAR" will be told to "go read the mailing list archives", "go read UPDATING" or simply "go away". With an XML-based dependency system in place, the blame would be entirely on the developer who had forgotten to update the NOTES file or had done so incorrectly. An XML-based list of supported devices and options would also allow the creation of a kernel configuration tool, similar to what Linux has had for = as long as I can remember. Again, a good thing for the end user. As it currently stands, FreeBSD is not designed for users who don't read the mailing lists to keep track of changes. Their opinions simply do not matt= er. AK