From owner-freebsd-current Tue Mar 21 07:03:32 1995 Return-Path: current-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id HAA21836 for current-outgoing; Tue, 21 Mar 1995 07:03:32 -0800 Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.34]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id HAA21822 for ; Tue, 21 Mar 1995 07:03:22 -0800 Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id BAA18525; Wed, 22 Mar 1995 01:01:42 +1000 Date: Wed, 22 Mar 1995 01:01:42 +1000 From: Bruce Evans Message-Id: <199503211501.BAA18525@godzilla.zeta.org.au> To: phk@ref.tfs.com, uhclem%nemesis@fw.ast.com Subject: Re: Can we create isa/matcd? Cc: current@FreeBSD.org, uhclem@nemes\is.lonestar.org Sender: current-owner@FreeBSD.org Precedence: bulk >Yes, I think it is a very sensible suggestion. Either I or Jordan will >work with you on this. I was hoping matcd could be changed to be more like the other drivers instead. I don't like big drivers such as isa/sound/* and isa/pcvt/*. matcd.c is inconsistent with most other drivers in its #include's. #include'ing "options.h" will probably break when someone fixes option handling in /usr/sbin/config, and #include'ing "creative.h" will break if anyone writes a driver named `creative'. See all the other drivers for examples (mostly bad :-]). The SCSI drivers somehow get by with not exporting any #defines. I like this. It helps to have an extra layer above the raw devices. The fd driver started right but now has ugly interfaces to the ft driver. The sio driver has the chip-dependent and bus-dependent header files more or less back to front. The syscons driver ignores all the keyboard #defines in ic/i8042.h and puts its own keyboard defines in the wrong place (). Bruce