RE: [Seamoby] CAR DiscoveryProtocol Requirements...

Hi Hesham,

comments inline

> 
> => To me the following text from the draft suggests
> that the AR knows ALL GAARs. I think that's unrealistic.
> 
>   4.1 Anticipated CAR Discovery
> 
>    In this approach, an AR currently serving the MN 
> identifies all CARs
>    for the MN's handoff, at some point prior to handoff.
> 
> 
>   > 
>   > It is obvious that CAR does not guarantee full knowledge, 
>   > but it has 
> 
> => So we agree, then I suggest you reflect it in the text.

As I outlined in my mail, the identification of CARs at the AR
is bound the current knowledge of GAARs at the AR. Out of this
set of GAARs, all CARs are identified. The text does not imply that
all GAARs are known.

> => You\re making a lot of assumptions about the discovery
> mechanism. If a MN can hear RAs from different
> GAARs, it will discover them. If that is not 
> possible due to l2 limitations, there are ways to go
> around that, without depending on any mobility
> management protocol, but keeping in line with
> existing protocols. 
> I didn\t want to start arguing about solutions but
> I think it can be done.

I don't understand your point here since I don't see the 'many
assumptions'
that I make. I'm trying to address requirements to solve the
problem. This is to identify GAARs under the constraint to minimize
the ARs that are actually not GAARs of the considered AR. I don't care
about MN involvement or not. This is not prohibited by the current set
of requirements, and I agree, it shouldn't. You should look at the basic
text of the requirements.

> => anticipation should not be associated with the 
> case where the MN does not signal, you can have
> anticipation when the MN signals also. Therefore
> associating 'anticipation' with a certain model
> where the MN does not signal, does not seem right.
> 
> I'm referring to the title of ch 4.1

I would appreciate if you could point out WHICH wording does imply
that MN signalling is ruled out in the anticipated approach. It is 
certainly not the title. 

Regards,

Dirk

_______________________________________________
Seamoby mailing list
Seamoby@ietf.org
https://www1.ietf.org/mailman/listinfo/seamoby