Comments of Antti Tuominen on CTP draft
"James Kempf" <[email protected]> Tue, 21 Oct 2003 14:44:30 -0700
| Newsgroups | gmane.ietf.seamoby |
|---|---|
| Message-ID | <02d701c3981c$84126490$0a6015ac@dclkempt40> |
Forwarded.
jak
---------------------------------------
I-D: draft-ietf-seamoby-ctp-04.txt
Review comments
---------------
1. In 2.3 it is specified that length is in 8 octet units. However,
if presence vector and data are not present we have 4 octet
message. Also context data may not 64bit align (especially when a
presence vector is used). So it seems to me that you either have
to have 1 octet as the unit (which of course limits length to 255
bytes) or do padding.
2. Also in 2.3 you use presence vector and presentation vector, so I
guess s/Presentation Vector/presence vector/.
3. In 2.4.1, what are Type=Auth-Token and Type Len? No description,
seem redundant. Same thing in 2.4.3 (also Algorithm) and 2.4.6.
Old remnants?
4. Example in Appendix C is wrong. First, you do not "subscribe" to
all-nodes. Node never sends MLD Report or Done for all-nodes.
Second, you don't put Routing Header in MLD messages, but Router
Alert Hop-by-hop option which is just 6 bytes (or 8 padded). In
consequence, all time and space overhead calculations are
incorrect.