Alvin:
 
I agree with you.  Personally, I like the simple differentiation between Compensable Jitter (CJ) and Non-Compensable Jitter (NCJ).  That's what I think SAS should use.  I believe those terms are the easiest, most straightforward, and least confusing way to communicate these principles. If it is desired, NCJ can be separated into its DJ part and its RJ part if such a distinction is needed.  As far as I know, the CJ/NCJ terminology shouldn't conflict with other pre-existing definitions.   An added attraction is that those terms are not new and are already defined as they have been discussed in this forum for some time now. 
 
Thanks for your feedback.
 
/Larry
 
-----Original Message-----
From: Alvin.Cox@seagate.com [mailto:Alvin.Cox@seagate.com]
Sent: Thursday, May 24, 2007 2:09 PM
To: Larry McMillan
Cc: t10@t10.org
Subject: Re: SAS PHY teleconference 5/24/07, 10 am CDT: Definition of BUJ and T10/07-259r0


I understand your concerns, however, rather than spend too much time on the terms to get to the end result, let's start from the end result and work backwards. I think the bottom line Mike was trying to share is:

TJ at ref RX output is "noncompensable jitter"

Instead of picking on the terms, how about proposing what you think SAS should use? I'm in no way intending to be disrespectful, but you are correct that the terms don't match up with these other references and we may need to define something new to accomplish what we want. Such a proposal will help us to complete the spec.

Thank you for the references.

Alvin Cox
Seagate Technology, LLC
Office 405-381-8067
Cell 405-206-4809
E-Mail  alvin.cox@seagate.com