Re: [DNA-BOF] Chairs' Charter discussion update.

[Greg Daley <greg.daley@eng.monash.edu.au>]

Hi JinHyeock,

JinHyeock Choi wrote:
[some cut]
> 
>>The second thing is that there is some interest
>>from one of the other Internet Area WG chairs
>>to improve the terminology (similarly to what
>>we have discussed here).
>>
>>Essentially, what was suggested was to discuss
>>'subnet change' instead of '(L3) link change'.
>>
>>I guess this would be based on existing RFCs which
>>say:
>>
>>(RFC3513 Section 2.1)
>>"Currently IPv6 continues the IPv4 model that a
>>  subnet prefix is associated with one link.
>>  Multiple subnet prefixes may be assigned to the
>>  same link."
> 
>  
> 
>>If this is acceptable to the AD's and the WG
>>chairs, would this be OK with the group?
> 
> 
> I checked the charter and think we can substitute 'link change' 
> with 'subnet change' (maybe with a little modification). It's O.K 
> for me to rephrase DNA charter with 'subnet change' instead of
> 'link change'. 
> 
> But do you propose to replace the term '(L3) link' with 'subnet'
> entirely? I think that 'subnet' and 'link' are different notions.  
>  
> According to 'Mobility Related Terminology (draft-ietf-seamoby
> -mobility-terminology-05.txt)', subnet is as below
>  
>      Subnet
> 
>        A subnet is a logical group of connected network nodes. In IP
>        networks, nodes in a subnet share a common network mask (in IPV4)
>        or a network prefix (in IPv6).
> 
> Hence if a router advertises two different prefixes, there may be two 
> different subnets in one link.  
>  
> The notion of 'link' is useful for DNA work, so I wish that we keep using it, 
> even though we rephrase DNA charter with 'subnet change.' 
> 
> Thanks for your work for furthering DNA. I think we are almost there. :-)

I agree that links and subnets aren't the same.

Regarding that, what we're really trying to do
in DNA is to determine changes to IP configuration.

IP configuration relies upon the reachability of a
subnet and its routers, not on the link itself.

So link change (without mutlilinks (sic)) implies
subnet change and subnet change implies address change.

Changing links implies changing router interfaces in
any case.

How about we just try to get the charter worked
out as unabmiguous (perhaps with L3 Links) and find an
agreement through discussion over the next couple of
weeks?

Greg