i think actually that if we just had an IDE controller, it wouldnt 
evolve, except maybe in the hands of one or two people on their own- i 
think it woudl be 'enough' and once a system was working, most people would
be less inclined to turn their system upside down a second time to make the
change to go MSCP. I think instead, the MSCP implementation should be 
done first since its the least likely part of the project to survive given
the existence of other parts of the project.. i'm not saying people wont 
finish things, just from my own experience something thats 'good enough' 
tends to stay in use way longer than planned, and the improvements to it
originally planned tend to get put on the back burner.. 
plus, as noted, loading an already slow CPU up with extra work is not
a good idea... 

On another note, what does anyone else think of starting from the supnik 
rqdx3 emulator? 

isildur

On Thu, 24 Jan 2002, David Woyciesjes wrote:
> 	Well, it would be a starting point for that intelligent controller,
> wouldn't it? Basically, build a basic IDE controller, and then once it's
> working solid, it can "evolve". I think it's best to start with a target
> that's a little closer and easier to hit...
>