RE: [mobile-ip] Re: FW: I-D ACTION:draft-tsirtsis-dsmip-problem-00.txt

[Pekka Savola <pekkas@netcore.fi>]

Sorry for a late reply..

On Thu, 31 Jul 2003, Tsirtsis George wrote:
> I am not sure where Mipv6/Ipv6/Mipv4/IPv4 comes from in your e-mail and I am
> not even sure what it means. MIPv4 and MIPv6 are signaling protocols....and
> do not introduce overhead in the data path.

That is incorrect.  MIPv6 and MIPv4 do introduce a data path overhead, in 
the form of the inclusion of IP-in-IP encapsulation, Routing headers and 
Home Address options (etc.).  I think this is what Alain was referring to.

> We can always complicate things but why not deal with the simple and by far
> most important issue.
> 
> Today Mobile IP signals HoA to CoA bindings of the same version resulting in
> either IPv4 over IPv4 encapsulation or IPv6 over IPv6 encapsulation.
> 
> All I am suggesting is that Mobile IP should be able to signal HoA to CoA
> bindings of different versions so that IPv4 over IPv6 encapsulation or IPv4
> over IPv6 encapsulation is also possible.

And what benefit, exactly, would the change in encapsulation have?  The 
Home Agents and the nodes would still have to support both versions, you 
would just end up with two Mobile IP protocols which (to some extent) 
supported both IPv4 and IPv6.

Looking briefly at the conclusions section gives me an impression you want
a transport-protocol independent MobileIP++ that's agnostic of IPv4 or
IPv6, with an understanding that you would not have to solve the problem
of the direct communication between IPv6-only and IPv4-only nodes.

It seems like HIP fits the bill quite nicely, and would be architecturally 
better approach than trying to glue MIPv4 and MIPv6 together somehow.

-- 
Pekka Savola                 "You each name yourselves king, yet the
Netcore Oy                    kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings