Re: [Autoconf] issues with draft-baccelli address model

[Alexandru Petrescu <alexandru.petrescu at gmail.com>]

Charles E. Perkins a écrit :
[...]
> So far the only hard problem seems to be how to avoid squabbling over
>  whether link-locals are needed.
>
> I am not sure why the proponents of necessity are so adept at 
> ignoring the solutions that do not need link-locals.

It's very simple why: I have a prototype which involves protocol
extensions (RA) only on the parts of the System which use exclusively
link-locals addresses.  You forbidding link-local addresses equates with
forbidding my protocol extensions on the core of my prototype, makes all 
break.

(the parts of the System which use global addresses are not affected by
my prototype.  They are also essential and core, but to some other people.)

> It is not explainable by the existence of solutions that use
> link-locals, because such solutions are universally acknowledged to
> be useful in their realm of applicability.

Is this formulation ok, or an "only" missing somewhere?

If you keep that formulation as is - I agree with it!

> So why do link-locals insist on "their way come hell or high water"?

BEcause I don't want my prototype discarded by an illogical sense of 
forbidding LLs.

Alex

> > See also Jari his last posting. This motivated me keep telling what
> >  I am doing and what the problems are. This also makes very clear
> > we shall stop boil the ocean.
>
> I agree with that.
>
> Regards, Charlie P.
>
> _______________________________________________ Autoconf mailing list
>  Autoconf at ietf.org https://www.ietf.org/mailman/listinfo/autoconf