I've read this a couple of times and I find the security section (sec 8)
quite confusing.  I am not a security expert but it appears to me that
it is not consistent.  

In particular sec 8.2 says "AH [RFC-2402] must be supported."  It then
goes on to say "there is no real need for AH" and in both section 8.1
and 8.3 there are items that "MUST be supported if AH is implemented".  
It would seem the if is redundant or something is wrong.

Equally, section 8.1 says that "IPSec tunnel mode MUST be supported" 
and then goes on to say "case .... MUST be supported if IPSec tunnel
mode is implemented."

The first paragraph of section 8.3 finishes with the sentance "Note that
the IPSec WG also recommends not using this algorithm."  It is not clear
to me which of the three algorithms mentioned in that paragraph this
sentance refers to.

It seems from section 8.3 that there are four encryption algorithms that
must be supported AES-128-CBC, HMAC-SHA-1-96, HMAC-MD5-96 and
HMAC-SHA-256.  I think this section could however be worded more
clearly.  It would also be good if the appropriate RFCs were referenced
in the text.  

>From the point of view of very small devices, whilst I understand that
IPSec support is a requirement, it seems that requiring transport mode
and tunnel mode, AH and ESP and four algorithms (plus null encryption)
seems onerous.  I wasn't part of any discussion on this, but I would
appreciate it if someone would explain particularly why so many
algorithms are required.

Finally a small editorial nit.  There are lots of "is MUST"s  and few
"is SHOULD"s in the document that should be "MUST"s and "SHOULD"s.

Richard.
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to