RE: RFC7084

Although I agree that asking address info whatever flag-settings, i.e. ignoring the flag, is not a bad approach to get things up and running from time-to-time.  On the other hand, it is a pitty that we just can't use them of course, i.e. use them where they are designed for.
Next to that, I don't agree that "trying to get an IP address (i.e. ia_na, not meaning ia_pd)"  all the time is often ok.  Reason is that the majority of our customer base is not numbering the WAN-link through ia_na nor slaac.  Although pd_excluce with ia_na is a nice extra thing to number them in a "clean" way.

Thx for the responses.  
Carl

-----Original Message-----
From: Ole Troan [mailto:otroan@employees.org] 
Sent: dinsdag 10 december 2013 16:14
To: Wuyts Carl
Cc: Simon Perreault; 6man WG
Subject: Re: RFC7084

Carl,

> Agree, but it does not mandate ia_na, does it ?

it is a hint, but wouldn't it be quite silly of an implementation to not try to get an address with whatever method works?
(if I implemented a CPE I would ignore the M/O flags).

cheers,
Ole

> -----Original Message-----
> From: Ole Troan [mailto:otroan@employees.org] 
> Sent: dinsdag 10 december 2013 15:59
> To: Wuyts Carl
> Cc: Simon Perreault; 6man WG
> Subject: Re: RFC7084
> 
> Carl,
> 
>> M=1 equals Managed flag = request ia_na or/and ia_pd.
>> Please note a router is also acting as host.
> 
> that is wrong. the M-flag does not give a hint about prefix delegation.
> the M flag is a hint from the network that stateful address assignment might be available.
> 
> cheers,
> Ole
> 
>> -----Original Message-----
>> From: Simon Perreault [mailto:simon.perreault@viagenie.ca] 
>> Sent: dinsdag 10 december 2013 15:28
>> To: Wuyts Carl; ipv6@ietf.org
>> Subject: Re: RFC7084
>> 
>> Le 2013-12-10 09:25, Wuyts Carl a écrit :
>>> M=1 should not be equal to force request ia_na.
>> 
>> What should M=1 mean then?
>> 
>>> what's the purpose of having separate options (ia_na and ia_pd)  if you're going to request both of them anyway ?
>> 
>> The answer to that seems simple to me: an end host would never request IA_PD. Only routers would.
>> 
>> Simon
>> --
>> DTN made easy, lean, and smart --> http://postellation.viagenie.ca
>> NAT64/DNS64 open-source        --> http://ecdysis.viagenie.ca
>> STUN/TURN server               --> http://numb.viagenie.ca
>> --------------------------------------------------------------------
>> IETF IPv6 working group mailing list
>> ipv6@ietf.org
>> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
>> --------------------------------------------------------------------