From owner-freebsd-hackers Thu Jun 4 13:59:52 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id NAA05854 for freebsd-hackers-outgoing; Thu, 4 Jun 1998 13:59:52 -0700 (PDT) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from dyson.iquest.net (dyson.iquest.net [198.70.144.127]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id NAA05629; Thu, 4 Jun 1998 13:58:55 -0700 (PDT) (envelope-from toor@dyson.iquest.net) Received: (from root@localhost) by dyson.iquest.net (8.8.8/8.8.8) id OAA02171; Thu, 4 Jun 1998 14:20:14 -0500 (EST) (envelope-from toor) Message-Id: <199806041920.OAA02171@dyson.iquest.net> Subject: Re: kernfs/procfs questions... In-Reply-To: <199806041910.NAA03447@mt.sri.com> from Nate Williams at "Jun 4, 98 01:10:58 pm" To: nate@mt.sri.com (Nate Williams) Date: Thu, 4 Jun 1998 14:20:14 -0500 (EST) Cc: dyson@FreeBSD.ORG, nate@mt.sri.com, mike@smith.net.au, hackers@FreeBSD.ORG From: "John S. Dyson" Reply-To: dyson@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Nate Williams said: > > And, I'm not stating that it's to be taken to the polar extreme either, > but that it's a *better* solution than sysctl. It's still not the best > solution either, but extending a poorer solution is certainly a step > in the wrong direction. > > I agree with Bruce in that programs are generally a better way of > configuring things. It's obvious if you know the system what needs to > be run, and how to get help on it. It also makes documenting things > easier, which sysctl does not. People already hate to document, and > making it hard to figure out where/how to document things just makes it > that much less likely to be documented. > I disagree for easily one simple reason: sysctl affords an internal documentation scheme that isn't a hack. Other reasons include the overhead and complexity of doing filesystems. If someone wants do do a kernfs that is as useful as our sysctl, and also figure out a clean way to provide the internal documentation, I wouldn't mind. Also, a general program that changes the binary representation provided by the kernel into readable text would also be important. This would require a kernel independent translation scheme. Right now, kernfs is too primitive. Also, it seems that the new kernfs shouldn't be optional. Sysctl as it is today, isn't optional, and we are very dependent on some of it for initial configuration (running a few times after startup sometimes.) We usually don't bang away at sysctl items, because they are often kernel parameters only. Sysctl provides a lot more than an initial view of it implies. Frankly, if we could do multi-stream files on /procfs, /kernfs, it would be great, because we could have seperate format fields and data fields. Too bad that doesn't exist yet. (Again, I find ASCII formatting in the kernel to be retrograde, and forgets about internationalization.) -- John | Never try to teach a pig to sing, dyson@freebsd.org | it just makes you look stupid, jdyson@nc.com | and it irritates the pig. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message