Thomas, >> We did have, apart from >> draft-schmidt-multimob-fmipv6-pfmipv6-multicast, several other >> multicast handover solution drafts such as: >> http://tools.ietf.org/html/draft-vonhugo-multimob-cxtp-extension-01 > > This draft basically uses the context transfer protocol (RFC4067) to > carry multicast state between ARs/MAGs. No full handover operations > have been specified.
It is because this draft inherits the basic operation from rfc4067 and hence no need to repeat to write it. The main aim of this draft is to define the message format for multicast context transfer with CXTP. Of course, if something is missing and required to clarify in the document, the authors will try to improve the document quality, I believe. > At the unicast side, RFC4067 is a predecessor of the more advanced > Fast Handover Protocols does not apply to unicast handover > management.... IMO it does not make much sense to consider this rather > elementary approach any further. > >> http://tools.ietf.org/html/draft-hui-multimob-fast-handover-04 >> > > This draft is - if you want so - a competitor to > draft-schmidt-multimob-fmipv6-pfmipv6-multicast, but has never been > worked out (as have several other attempts in the past). If this > document was to be advanced, it had to rewrite (or copy ??) 80 % of > our draft, which is not a proper way to treat authorship. The -00 version of draft-hui was published on June 29, 2009. Your -00 draft was published on March 01, 2010. (I just tell the fact.) >> If we decide to accept more than one handover solution then we >> probably need to consider all of them for possible WG adoption. > > I don't understand your thinking here: it is perfectly normal that > there are competing approaches and the idea of the IETF discussion is > to have the best solution win. From the performance side, and from > protocol systematics (cooperation with unicast), > draft-schmidt-multimob-fmipv6-pfmipv6-multicast is clearly the best > solution for fast handover operations - this comparison includes the > adopted draft "fast handover from transient binding". If only one HO document is needed, then since this WG has already been working for SIAL proposal, other documents (including yours) related to HO are not needed, are they? Furthermore I don't think all mobile operators in the world adpot only FPMIPv6 as their business. So, if other idea is useful and technically correct, documenting it is reasonable. This is just my feeling. Regards, -- Hitoshi Asaeda _______________________________________________ multimob mailing list [email protected] https://www.ietf.org/mailman/listinfo/multimob
