Date: Tue, 21 Oct 1997 09:46:40 +0100 (MET) From: Luigi Rizzo <luigi@labinfo.iet.unipi.it> To: mike@smith.net.au (Mike Smith) Cc: tg@ihf.rwth-aachen.de, mark@vmunix.com, freebsd-stable@FreeBSD.ORG Subject: Re: PNP and new Sound code ? Message-ID: <199710210846.JAA20825@labinfo.iet.unipi.it> In-Reply-To: <199710210850.SAA00578@word.smith.net.au> from "Mike Smith" at Oct 21, 97 06:20:18 pm
next in thread | previous in thread | raw e-mail | index | archive | help
> > No. I hope it's in xperimnt (or whatever this is called) and I'd > > really like to bring at least the PnP support over to 2.2-STABLE after > > the release (I don't care for sound). > > I don't think that would be a smart idea at all. The most recent > changes appear to have broken it on -current, and the code is certainly > far far away from being ready for a production-level system. [WARNING: People, don't thing this is a personal debate with Mike, whose work and opinion I highly respect and I am sure the same holds for him. I am just taking this chance for showing my position about -STABLE] May I ask what kind of brokennes you are talking about ? If you refer to the message that Jordan sent out some time ago, he promptly took it back saying that it had nothing to do with the cancelled 3.0 snap. I am well aware that my PnP code, as it is now, is not more than a manual configuration utility for ISA-PnP cards. It does not want to do more, by design and lack of energy on my side. In particular, it does not want to be a solution to the problem of resource allocation. This does not mean that it is unusable (maybe with minor patches, or a little bit more of documentation, if the one that i have already written is not enough) on a production system. We have many many pieces of code in the source tree, even in -stable, which are far away from being optimal solutions to the problems they should solve. If we went for the "optimal or nothing" road, we would not have support for many devices which are now in (ATAPI CD, Voxware 3.0 sound, gpib, ...), we would not have many broken pieces of code inherited from 4.4 or older version of FreeBSD, we would not have hacks which use the "flags" field in the kernel config file, etc. etc. Really I am not particularly concerned with the PnP stuff, but at times have the feeling that some positions are too conservative. I think similar positions have come out over the years regarding other things, such as ATAPI CD, or IDE disks, or the "de" driver, and I have been always disappointed in seeing these discussions. There is a market we should look at, not perhaps as a first goal, but we cannot ignore it. If there are many PnP devices around, we should at least try to support them, or people will migrate elsewhere (to other OS, or to other hardware). Just as a data point. During the last three months I have tried to support both the PnP and the audio driver. Of the about 500 messages I have saved in my mailbox, only a negligible fraction of them refers to PnP problems, which means that either people did not have problem with that code, or did not use it (but i doubt that my audio driver even compiles without the PnP code). Cheers Luigi -----------------------------+-------------------------------------- Luigi Rizzo | Dip. di Ingegneria dell'Informazione email: luigi@iet.unipi.it | Universita' di Pisa tel: +39-50-568533 | via Diotisalvi 2, 56126 PISA (Italy) fax: +39-50-568522 | http://www.iet.unipi.it/~luigi/ _____________________________|______________________________________
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199710210846.JAA20825>