Re: [rfc2462bis] whether we need the M/O flags

[Alain Durand <Alain.Durand@xxxxxxx>]

Jinmei,

I agree with your analysis, except that I'm not so much worried about
breaking somebody's assumption in designing an hypothetical client
dealing with the O/M bits. The definition of those bits was obscure
in 2462 anyway.
I would be more worried if this was to result in demonstrated 
operational
problems. I honestly believe that there won't be any as deprecating O & 
M
will end up being equivalent to today's most common practice where
those bits are not set. Nothing breaks today under those circumstances,
so I do not see what should break tomorrow if we were to deprecate 
those 2 bits!

Of course, if someone could point to a real case scenario where
this would be a problem, that could change my opinion...

	- Alain.


--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------