Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1030254AbWAGAk4 (ORCPT ); Fri, 6 Jan 2006 19:40:56 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1030247AbWAGAk4 (ORCPT ); Fri, 6 Jan 2006 19:40:56 -0500 Received: from ds01.webmacher.de ([213.239.192.226]:23233 "EHLO ds01.webmacher.de") by vger.kernel.org with ESMTP id S1030245AbWAGAkz (ORCPT ); Fri, 6 Jan 2006 19:40:55 -0500 In-Reply-To: References: <20060103215654.GH3831@stusta.de> <9a8748490601031411p17d4417fyffbfee00ca85ac82@mail.gmail.com> <0D76E9E1-7FB0-41FD-8FAC-E4B3C6E9C902@dalecki.de> <1136486021.31583.26.camel@mindpipe> <20060106034026.c37c1ed9.diegocg@gmail.com> <20060106145723.GA73361@dspnet.fr.eu.org> Mime-Version: 1.0 (Apple Message framework v746.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Message-Id: <1CE30648-5101-4A2D-A240-64381770A9A6@dalecki.de> Cc: Olivier Galibert , Diego Calleja , rlrevell@joe-job.com, jengelh@linux01.gwdg.de, Takashi Iwai , jesper.juhl@gmail.com, bunk@stusta.de, zdzichu@irc.pl, s0348365@sms.ed.ac.uk, ak@suse.de, ALSA development , James@superbug.demon.co.uk, sailer@ife.ee.ethz.ch, linux-sound@vger.kernel.org, zab@zabbo.net, kyle@parisc-linux.org, parisc-linux@lists.parisc-linux.org, jgarzik@pobox.com, linux@thorsten-knabe.de, zwane@commfireservices.com, zaitcev@yahoo.com, LKML Content-Transfer-Encoding: 7bit From: Marcin Dalecki Subject: Re: [2.6 patch] schedule obsolete OSS drivers for removal Date: Sat, 7 Jan 2006 01:40:27 +0100 To: Jaroslav Kysela X-Mailer: Apple Mail (2.746.2) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2791 Lines: 65 On 2006-01-06, at 21:26, Jaroslav Kysela wrote: > > I got the point, but the audio is very specific that it requires > realtime > scheduling. Even graphics is not so tied with realtime, because a > scheduling gap does not end with error or very ugly misbehaviour > (pops) > like in audio. Nonsense. You just got too used to utterly bad behavior imposed by the inherently defective X11 graphics system. Like stuttering mouse movements. Like race conditions between popup menus and background menus and so on and so forth... Hick-ups when suddenly the system starts to do some paging... More staggering examples simply don't occur that obviously in praxis because the interface designers refrain from using some effects like subtly animated icon change and menu animations redrawing. But it's by no way an argument which could be used to justify the deficiency of some audio system. And BTW. A proper user interface system requires synchronization between audio and video. So there you are - all over the pill - soft real time requirements. Actually not that soft at all. It always surprises me how efficient and demanding the perceptive system of a hunter and gatherer is, which only just recently got the sudden idea that it may be nice to spend his time in front of an engine generating animated images. >> At least OSS, with all its flaws, is a documented kernel interface. >> You can static link a oss-using program, whether it uses it directly >> or through interfaces like sdl-audio, and it will just work. > > Please, see your words. You're simply anarchistic. You replaced > flexibility of dynamic library with a possibility to have static > binary. > Also note, that if OSS had the API in userspace from the first days, > the emulation or redirection of this API to another API or user space > drivers wouldn't be so much complicated nowadays. Bummer. You simply don't get it. On unix like systems the definitive interface level is not a library. It's the system call. Libraries can be helpful as a means to make some common actions a tad bit easier. However even so simple things like state- full behavior there is a complete no go. Get over it. Libraries are just too affine to compiler releases particular programming languages and many other side conditions. Do your homework and take a serious look at other operating systems to see how "fine" the APIs defined by DLLs (ups) are. As an exercise ask yourself why on earth it takes usually about 2 or 4 times longer to write a windows driver. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/