We already have a 'pciide' vs 'wdc' attachment.
	Why not have 'pciide', 'ide', and 'wdc' attachments, allowing the
	'ide' driver to keep up with the latest IDE standards and keeping
	all the ESDI/MFM legacy code separate. The 'ide' driver can probe
	and attach first. If it fails, the 'wdc' driver could try.
	
	If they both could attach, then pick the one which matches the
	most drives :)

		David/absolute

On Thu, 16 Dec 1999, Brett Lymn wrote:

> According to Michael Graff:
> >
> >Can the driver autodetect ESDI vs. IDE, or perhaps use a compile-time
> >option?
> >
> 
> I don't know but I think that it would be desirable if this could be
> done.  Todd's suggestion of having a "legacy" kernel may be a good way
> out.
> 
> >I know the whole point is to run on anything, but I would hope the
> >default is to run well on things made recently.
> >
> 
> Yes this is the correct thing to do but handling old devices
> gracefully is a cool thing to do as well, especially if it does not
> come at the expense of the new devices.  I can see that we will only
> lose if we tell people "that is crap old hardware - get a new drive"
> when all they are trying to do is lash up a router (or similar) on the
> cheap and all they could lay their hands on was an ESDI drive in a
> 486.  I know drives are cheap but nothing beats the price of
> "under-spec, old" hardware that has been rescued from the dumpster ;-)
> 
> 
> -- 
> ===============================================================================
> Brett Lymn, Computer Systems Administrator, British Aerospace Australia
> ===============================================================================
> 
>