Possibly more importantly, this is one of the best chances for the *NIX
community to put aside the petty bickering and infighting that has spoilt
its' chances in the past. It's easier to get a SCO user to switch to
Linux, than to get an NT user to make the switch. If they know that the
exact same hardware that runs under SCO will also work with Linux, that's
one less reason to put them off.
The stability of closed drivers is an issue, as is what sort of work will
be required to get the kernel to support UDI. If it means doing things in
some bass-ackward way, perhaps UDI is best left alone. But if it can be
done cleanly, it probably should be done. After all, until Adaptec opened
up, we supported the 2940 and such as best we could, but said 'buy
Buslogic if you possibly can'. As has already been pointed out, it may
be possible to implement the UDI framework such that it can give useful
debugging info about what a UDI driver is doing, which would help both
with pointing the finger at faulty OEM UDI drivers and with reverse
engineering a native driver.
Finally, my reading of the press release implied nothing about porting
Linux drivers to the UDI model. Intel's intention is to help us develop a
UDI framework for OEM drivers.
My 2 farthings,
Alex Butcher.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/