From owner-freebsd-current@FreeBSD.ORG Thu Jan 10 15:41:30 2008 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 A337316A418; Thu, 10 Jan 2008 15:41:30 +0000 (UTC) (envelope-from rwatson@FreeBSD.org) Received: from cyrus.watson.org (cyrus.watson.org [209.31.154.42]) by mx1.freebsd.org (Postfix) with ESMTP id 7961713C457; Thu, 10 Jan 2008 15:41:30 +0000 (UTC) (envelope-from rwatson@FreeBSD.org) Received: from fledge.watson.org (fledge.watson.org [209.31.154.41]) by cyrus.watson.org (Postfix) with ESMTP id A81D047727; Thu, 10 Jan 2008 10:41:29 -0500 (EST) Date: Thu, 10 Jan 2008 15:41:29 +0000 (GMT) From: Robert Watson X-X-Sender: robert@fledge.watson.org To: Adrian Chadd In-Reply-To: Message-ID: <20080110153814.C2499@fledge.watson.org> References: <478556AD.6090400@bsdforen.de> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: Dominic Fandrey , freebsd-current@freebsd.org Subject: Re: FreeBSD's problems as seen by the BSDForen.de community 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, 10 Jan 2008 15:41:30 -0000 On Thu, 10 Jan 2008, Adrian Chadd wrote: > Again, if someone there wrote up some performance test suites or built a > testing environment then I'm sure you'll get some help. > > The FreeBSD has always been helpful when people willing to develop have > stood up and engaged the community in coming to a solution. So, are you guys > willing to stand up and contribute something towards solutions to the > problems you're seeing? Indeed -- experience suggests that we tend to do the best job at improving the performance and quality of code when there are willing hands with quantitative measures by which to evaluate progress. Something I'd really like to see, for example, are some tests that look at the latency of operations visible to users in the windowing system and reduce the results down to a set of simple values that can be used to drive optimization. This has worked very well with workloads like SQL processing, DNS, etc, and I think could work well in other areas. The key here is getting people who care about workloads to invest time to try and make them (or comparable workloads) accessible to developers. Coming up with those benchmarks, or at least measures, proves tricky and requires quite a bit of technical expertise, but I think it's often more accessible than coming up with the fixes and architectural changes required to improve numbers. But it's surpising how motivation a simple set of performance numbers can be in the right hands. Robert N M Watson Computer Laboratory University of Cambridge