robertl@killer.UUCP (Robert Lord) (07/30/87)
I was talking to a friend the other day, and he said he worked for a
Company called Q-1 which had made a disk controler which would calculate
the distance that it had to jump to get to the track it needed to get to
(make sence so far?) and then would put the head into hyperspace to jump to
that track. It usually came within 1 or 2 tracks of where it wanted to
be, so it just had to jump a track or 2. It was programed to expect errors
on the way there though....2 questions:
1) Is something like this still around? This combined with RLL
would make a very good combination!
2) How did they get the head to go faster?
Robert Lordayac071@ut-ngp.UUCP (08/01/87)
In article <1246@killer.UUCP> robertl@killer.UUCP (Robert Lord) writes: >I was talking to a friend the other day, and he said he worked for a >Company called Q-1 which had made a disk controler which would calculate >the distance that it had to jump to get to the track it needed to get to >(make sence so far?) and then would put the head into hyperspace to jump to >that track. It usually came within 1 or 2 tracks of where it wanted to >be, so it just had to jump a track or 2. It was programed to expect errors >on the way there though....2 questions: > > 1) Is something like this still around? This combined with RLL > would make a very good combination! > > 2) How did they get the head to go faster? I would be wary of such a device if I were you. I had one earlier, but didn't like it. Seems that the head would jump to hyperspace, then reappear at the correct track, but at the wrong time. Sometimes it would read data I had erased days ago, but occassionally it would read code I hadn't even written yet (I think nothing is worse than to see examples of code you have yet to write. I still shutter to think about it!) Anyway, until they fix the problem with time-space relativity, I wouldn't recommend these babys to anyone. BTW, does anybody want to make an offer on a used TARDIS? Slightly used, but in excellent repair. ;-) Bill Douglass ayac071@ngp.UUCP
john@hpcvlo.HP.COM (John Eaton) (08/03/87)
<<<<< < < 2) How did they get the head to go faster? < Head speed is usually expressed in the number of milliseconds that it takes to move the head one track. This number is highly dependent on the current head speed but is normally given at the worst case of moving a head that is complety stopped. The trick is that if you are moving the head over multiple tracks you can step it at a faster rate once you have gotten it moving. Smart controllers will do this to save time on long seeks. John Eaton !hplabs!hp-pcd!john