On Mon, 17 Jul 2006 14:34:09 +0200 Richard Braun <[EMAIL PROTECTED]> wrote:
> On Mon, Jul 17, 2006 at 02:35:51PM +0300, Constantine Kousoulos wrote: > > In that way, > > writing glue code is not very reusable. I could try writing > > gnumach native drivers, although i'm not sure this is a better > > approach. > > I disagree. Writing native drivers would take a *lot* of time. Just > consider how hard debugging a driver is. As Thomas suggested, we > are not forced to use Linux drivers. There are other alternatives out > there. I think I'll try FreeBSD 6 drivers first, and if I can get IDE > support working, it would be interesting to consider this approach. > BSD kernels are more or less known to have more or less stable kernel > interfaces, at least far more stable than in Linux. > FreeBSD drivers are a bit closer to Mach ones. A new framework to drivers on Mach would be a good idea. It would be closer to FreeBSD device driver framework, so, we could port FreeBSD drivers to GNU Mach. > > I could focus on porting linux's sound drivers as a primary > > develpment target. I have to admit though, that i was thinking on > > focusing on usb support first. > > Well, we should first make sure that there will be as few regressions > as possible if we switch to another drivers set. So getting IDE working, > then Ethernet NICs, and SCSI controllers already supported must be > the priority (IMHO). If the code base is correctly integrated, it will > not be particularly hard to add other driver classes in after that. > > > However, I would like to inform you that for the next 7 days i > > will take a summer vacation. I will continue the development when > > i return. > > Sure, nice holidays then :-). > > -- > Richard Braun > _______________________________________________ Bug-hurd mailing list Bug-hurd@gnu.org http://lists.gnu.org/mailman/listinfo/bug-hurd
