Re: Project UDI (Device Drivers)

[Pedro Giffuni <giffunip@xxxxxxxx>]

ibotty wrote:

>  The huge amount of work they made is explained on:
> >     http://www.projectudi.org/
> > and an initial port for FreeBSD 4.x can be found here:
> >     http://projectudi.sourceforge.net/
>
> both pages look dead. is this effort still alive?
> (prjectudi.sf.net only having one single newsitem for the last 2 years)
>

I did get some private mail from them like a month ago, about some interest to
get more BSD hackers so
they are not really dead. The specific device driver development does seem to be
mostly closed source.

>
> even if yes, it is still questionable, if we get any advantage on getting
> udi-support in the os. as far as i read on above pages, there are virtually
> no drivers.

I wrote that on the original email, as I said it's just FYI. They did do a lot
of work though (I bet you haven't
read the specs ;) )  so maybe their ideas have some space in an evolved OS like
DragonflyBSD, at least as
reference. They claim the real value in the framework is technical, not just
source and binary portability,  and they did a lot of benchmarking on their
framework.

>
> the linux camp will NOT switch to udi (or even develop anything for udi), so
> we are left with drivers, that are either closed-source by sco (which is in
> progress of dying), or which are so trivial, that we can develop them in a
> fraction of time, than finish the udi-wrapper.
>

The death or life of SCO and the success or failure of it's legal actions
lawsuits against IBM is not something that I care about ;). About the triviality
of developing device drivers, well... UDI is less than a week away of porting
effort  to someone that knows his way around BSD. I guess it's a matter of
determining if it's a better deal for a driver developer to find a sponsor for a
BSD driver or a sponsor for an UDI driver (I have NO idea).


>
> this is just my humble opinion, though.
> if someone wants to start, go ahead!
>

ditto.

    Pedro.