RE: On the NECESSITY of MIP for LMM (was Re: [Mipshop] WG Last Call comments...)

I will be sending my review I guess to multi6 of MAST.  This has nothing
to do with MIP. It has to do with implementation as HIPI and I think
crosses an IETF stadards line into implementation.  IETF don't tell
implementors how to build their stack.  Regardless what has MAST or HIPI
have to do with this discussion from your perspective?

thanks
/jim

> -----Original Message-----
> From: mipshop-admin@ietf.org [mailto:mipshop-admin@ietf.org] 
> On Behalf Of Pekka Nikander
> Sent: Friday, October 31, 2003 5:01 AM
> To: carlw@mcsr-labs.org
> Cc: jeanmichel.combes@francetelecom.com; gab@sun.com; mipshop@ietf.org
> Subject: Re: On the NECESSITY of MIP for LMM (was Re: 
> [Mipshop] WG Last Call comments...)
> 
> 
> Carl and Jean-Michel,
> 
> Just to clarify:  I am not asking the requirements
> draft to be extended to consider sub-IP mobility or
> other micro-mobility schemes.
> 
> I am still talking about plain, IP layer mobility.
> My point is that Mobile IP is only one IP layer mobility 
> method, admittedly the most mature one today, but not the only one.
> 
> What I have been asking for is to *possibly* rewrite
> the requirements document so that it would not mandate
> MIP based solutions.  The main motivation is to allow
> recycling of the requirements document for other work,
> for *example*, MAST or HIP based LMM.  A related but
> lesser motivation is that maybe we could later get IP
> level LMM schemes that are not based on MIP but better
> suited for the job at hand.  However, I am very well
> aware that such work will not happen within the
> current charter of this working group.
> 
> Anyway, I do definitely want the LMM requirements document
> to say that the solution MUST be compatible with MIP, and
> that the solutions considered MUST work at the IP layer
> (and not below it).  The *only* thing that I do not like
> is the requirements document to be solution specific from
> the outset.  That limits its scope unnecessarily.
> 
> [Nit picking:  The keyword here is "necessity".  What
> I tried to argue for is that there are no technical
> reasons that necessitate the requirements document to say
> that the solution MUST be based on MIP.  (The charter says 
> otherwise, and I am OK with the charter.)  Therefore it is 
> *unnecessary* to state that the solutions MUST be based on 
> MIP.  I argued that it would be trivial to rewrite the 
> document so that it allows other solutions. I still aim to do 
> that, time permitting.]
> 
> > I think (my opinion) is that we should move forward with
> > the LMM requirements draft as is (given editorial changes 
> suggested so 
> > far).  My concern is that a general requirements draft isn't as 
> > straight forward as is being stated.  The goal here is in advancing 
> > LMM in the context of mipshop
> > - which has a charter based on MIP solutions.
> > 
> > But I am open to working with Pekka on generic requirements 
> document 
> > if that is what the decision is.
> 
> Very much now depends on the understanding of "generic".
> To me it is generic in the sense that it does not limit 
> solutions to Mobile IP, but still works at the IP layer and 
> not by going below it.  MAST, HIP, LIN6 etc all work, in this 
> sense, at the IP layer.
> 
> --Pekka Nikander
> 
> 
> 
> 
> _______________________________________________
> Mipshop mailing list
> Mipshop@ietf.org
> https://www1.ietf.org/mailman/listinfo/mipshop
> 

_______________________________________________
Mipshop mailing list
Mipshop@ietf.org
https://www1.ietf.org/mailman/listinfo/mipshop