[OpenAFS-devel] Re: IPv6 strawman plan: Re: Moving Forwards

[Andrew Deason]

On Mon, 10 Sep 2012 12:29:46 -0400
Garrett Wollman <wollman@csail.mit.edu> wrote:

> >> Aside from what Simon said, anyone can just request RPC code points
> >> without explanation of what they are. If Troy wants to request 50
> >> 'private' points from each relevant package and make his own
> >> protocol, I have a hard time seeing the trouble with that.
> 
> > Requesting RPC code points would be following the standardization
> > and registration process.
> 
> No, that's the wrong way around.
> 
> A working implementation comes first, then you actually have
> sufficient information to consider standardizing it.  There is
> absolutely no value in standardizing something that has never been
> implemented.

Well, first of all, you can't have an implementation without allocated
code points.

But anyway, what I was trying to talk about is a case where a particular
protocol is never standardized (at least, not intended to be, at least,
not in the short term). You can request a bunch of code points for
'private/experimental' use, and then never tell anyone what they do.

So, Jeff is just saying, yes, that is following the "process". As in,
that is playing by the rules.

-- 
Andrew Deason
adeason@sinenomine.net