Re: iSCSI: Serial Number Arithmetic

["THALER,PAT (A-Roseville,ex1)" <pat_thaler@agilent.com>]

Julian,
 
The draft says serial number arithmetic as defined in RFC 1982 is to be used
for comparisons and arithmetic on CmdSN but but it uses operations which are
inconsistant with that statement.
 
2.2.2.1 (bottom half of page 26) re: the processing for MaxCmdSN and
ExpCmdSN - "and with a difference bounded by 2**31-1" should be deleted (3
places). The RFC 1982 Serial Number Arithmetic rules cover the meaning of
addition of a positive integer of limited range and comparison (<, > and =)
for serial numbers. "Only two operations are defined upon serial numbers,
addition of a positive integer of limited range, and comparison with another
serial number."  The operation difference is not defined for Serial
Numbers.<?xml:namespace prefix = o ns =
"urn:schemas-microsoft-com:office:office" />

2.2.2.2 re: "difference is greater than 2**31-1" - The operation difference
is not defined for Serial number arithmetic. Even if it was defined as the
inverse of addition, addition is only defined for adding a value up to
2**31-1. Is the intent of the statement that it should cover only the value
where the results of StatSN > ExpStatSN and StatSN < ExpStatSN are undefined
or is the intent that it cover that value plus the values where ExpStat >
StatSN? The former doesn't seem useful and the latter seems rather severe.

Regards,
Pat Thaler