RFC 3779, "X.509 Extensions for IP Addresses and AS Identifiers", June 2004Source of RFC: pkix (sec)
Errata ID: 2537
Publication Format(s) : TEXT
Reported By: Jim Schaad
Date Reported: 2010-09-30
Verifier Name: Sean Turner
Date Verified: 2011-02-23
Section 18.104.22.168 says:
To simplify the comparison of IP address blocks when performing certification path validation, a maximum IP address MUST contain at least one bit whose value is 1, i.e., the subsequent octets may not be omitted nor all zero.
It should say:
Text should be deleted.
There are a number of different issues relative to this text that need to be addressed.
1. This text implicitly change the rules for encoding a maximum value. As an example the address 0.0.0.255 is encoded as 03 03 00 00 00 00 according to the rule " The BIT STRING for the maximum address results from removing all the least-significant one-bits from the maximum address."
2. The rule in no way simplifies any comparisions of IP address blocks. If one really wishes to simplify the comparison then one needs to change the rule for maximum addresses to remove all but the last least-signficant one-bit from the address. However it is not clear that even this would really simplify the comparison in any significant way.
If you look at the example in 22.214.171.124 - tis is not clear how having the one bit at the top of the encoding helps make the comparisons any easier - but it satisfies the requirment that atleast one bit is a 1. If the maximum value ws encoded as 1000 1 (0x3 0x02 0x03 0x84) - a bitwise comparision routine could make for a simplified a < b comparison (looking at only the top 5 bits of the address to be compared.)