From owner-freebsd-current@FreeBSD.ORG Thu Sep 17 19:11:37 2009 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 616041065695 for ; Thu, 17 Sep 2009 19:11:37 +0000 (UTC) (envelope-from dougb@FreeBSD.org) Received: from mail2.fluidhosting.com (mx21.fluidhosting.com [204.14.89.4]) by mx1.freebsd.org (Postfix) with ESMTP id 120338FC1B for ; Thu, 17 Sep 2009 19:11:37 +0000 (UTC) Received: (qmail 23465 invoked by uid 399); 17 Sep 2009 19:11:33 -0000 Received: from localhost (HELO foreign.dougb.net) (dougb@dougbarton.us@127.0.0.1) by localhost with ESMTPAM; 17 Sep 2009 19:11:33 -0000 X-Originating-IP: 127.0.0.1 X-Sender: dougb@dougbarton.us Message-ID: <4AB289DF.5020303@FreeBSD.org> Date: Thu, 17 Sep 2009 12:11:27 -0700 From: Doug Barton Organization: http://www.FreeBSD.org/ User-Agent: Thunderbird 2.0.0.23 (X11/20090822) MIME-Version: 1.0 To: Marcel Moolenaar References: <200909010931.16880.nick@van-laarhoven.org> <1251841416.1689.4458.camel@balrog.2hip.net> <200909021656.15747.nick@van-laarhoven.org> <20090904100847.GA13167@server.vk2pj.dyndns.org> <20090904101630.GA17207@camelot.theinternet.com.au> <20090904135252.GA23438@lonesome.com> <642B63E0-0F75-45FD-9E8D-58D990F7C8EB@mac.com> In-Reply-To: <642B63E0-0F75-45FD-9E8D-58D990F7C8EB@mac.com> X-Enigmail-Version: 0.96.0 OpenPGP: id=D5B2F0FB Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: Mark Linimon , FreeBSD CURRENT Mailing List , peterjeremy@acm.org Subject: Re: Reducing noise in dmesg output X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 17 Sep 2009 19:11:37 -0000 Marcel Moolenaar wrote: > > On Sep 4, 2009, at 6:52 AM, Mark Linimon wrote: > >> No one has mentioned the other reason to leave in verbosity: so that >> users >> who are having problems can file more useful PRs. This is particularly >> true of video cards (which, as one might recall, is where this thread >> started.) > > This has always been an interesting fault-line. > > Yes, if you print or log "everything" then there's bound to > be useful information somewhere that can be used to analyze > problems. Approaching this from the glass half-empty angle, > I can see why people value verbosity. It's an easy case to > state: without it we don't know what went wrong. > > There's a flip-side and it's one that's much harder to argue > for. Arguments against verbosity include such things as: > 1. The signal/noise ratio is worse which means that it's > easier to miss the information that is truly important. > 2. You present the user with output that's not even directed > towards the user -- it's an aesthetic bug. > 3. It introduces performance problems, especially on slow > consoles. > 4. If it works, it works and the verbosity is unnecessary. > > Much more subjective... > > As long as we depend on verbosity to provide us with the > information we need to solve a problem, it's really hard to > convince people that we should make it more user-oriented > and print only things that are of value to the user. Which > means that unless developers value the user perspective and > are willing to put in the effort to allow for another way > of obtaining the information, verbosity is hard to reduce. > It's not in the developer's interest. That is, unless the > problem reporting is actually much better if done differently. Marcel has several good points here. I'd also like to point out that there is a middle ground which I haven't seen mentioned yet, dramatically reduce what goes to the console while sending all of what we do now (or potentially even more) to syslog by default. Doug -- This .signature sanitized for your protection