Doug,

I regret to say there's a problem with our ACM Conf.
paper (on disk cylinder remapping).  Last week we sent
a camera-ready version to Cherri at Auburn, and then,
over the weekend, we discovered an error.  The correct
performance improvement numbers for the real system
tests are different from those we reported.  Our 
proposed scheme still offers the best performance
(in fact, better than what we had reported), but the
pipe-organ approach also offers better performance than
we reported, and our relative win (over pipe-organ) is
not nearly as great.  

So, I see three options here:

1. Issue errata at the conference.
2. Send Cherri a revised camera-ready copy now.
3. Withdraw the paper.

I guess I favor 2, but it's really your call.

I'm sorry about this.  There were many graduate students
involved, and I simply didn't have time to check all
their work.  Let me know what you'd like us to do.

Regards,
Robert