Jason Wright drunkenly mumbled...
>
> As soon as someone can convince the marketriods at NPI that releasing
> their driver doesn't do them any harm development can begin.  It started
> the process almost a year ago (if not longer) and they were having none
> of it.

i'm trying to get uhm, whatever their name is now (used to be Artecon) to get
off the specs for the ArtePort card that is no longer supported in Solaris so
that i can use it in NetBSD.  this is a product that hasn't been made in years.
they don't sell them anymore, and interstingly enough, they aren't even in the
market share the ArtePort is a member of anymore, that make wacky SAN and NAS
hardware now.  so even if i stole the designs for the ArtePort and started 
building them it wouldn't make a difference to Artecon.  will they release any 
info?  of course not, that would been far too cool of them.

> I know what chips are on it, but most important thing though is the
> register mapping and then any hardware strangeness.  The hard part
> about writing supports for these cards in particular is that they
> do not handle ring negotiation internally; it must be done in
> -ick- software, so that means writing ring handling, too.

and who knows what wackyness exsists on the things to thwart someone like you
from trying to "reverse engineer" it through software.  maybe none, but i
wouldn't be supprised if there were.

this is all a little out of my area, but is there an advantage to doing ring
negotiation in software (other than relieving the firmware coders the extra
hassle, not that that's what they are paid for or anythgin)

> If someone can convince them to give me documentation and/or SunOS
> source I would be glad to write a driver as I have a ton of them
> (well, at least several) and I've got experience writing Sbus
> network drivers for BSD.

fat chance.  if i can't get Artecon to come up off the specs for the ArtePort
what chances do we have with NPI?

this sucks,

-brian