Re: Comments on NHRP spec.,, modes of deployment etc.

[Bruce Cole <[e-mail]>]

> > From [e-mail] Tue Aug 15 13:49:12 1995
> > To: Grenville Armitage <[e-mail]>
> > Cc: Bruce Cole <[e-mail]>om>, [e-mail], [e-mail]
> > Subject: Re: Comments on NHRP spec.,, modes of deployment etc. 
> > In-Reply-To: Your message of "Tue, 15 Aug 1995 10:51:27 EDT."
> >              <[e-mail]> 
> > Date: Tue, 15 Aug 1995 13:28:50 -0700
> > From: Bruce Cole <[e-mail]>
> 
> > (The ability
> > of 'hosts' to answer NHRP requests is still required since it was not
> > made a requirement that NHSes assume the full burden of providing
> > NHRP responses for stations that attempt to register with them).
> 
> Can anyone explain the reasoning behind this as a design choice and also
> point us to the relevent text in the document?

I do not know whether or not this was a design choice.  My reasoning was that
since the current spec does not require NHRP responses to be generated
before the packets reach end stations, one should not assume otherwise.
Remember, it was proposed that we get rid of server mode, so if stations
within the NBMA network do not exactly agree on who their neighboring NHSs
are, requests may be forwarded on to end stations which consider themselves to
only be clients.

Requiring an NHRP client to be able to answer NHRP requests of itself does
not seem unreasonable to me; ARP implementations have no problem with this.
But if the group wants to limit the complexity required in client 
implementations, then that's fine too.  Somebody please draft some text
to make this explicit.

Note that I believe Rob Coltun is drafting text to support the opposite -
where the NHRP requests can be required to go to the end stations.  This is
to better support multi-homed end stations.

> Is this yet another case where the NHRP authors are making some implicit
> assumptions without sharing them with the rest of the world so we can all
> implement the protocol too?

I'm sure that's not the intent, and I honestly don't know what other
cases you're eluding too.  Much of the text of the current draft has
simply evolved, perhaps at the expense of clarity.  Maybe you could submit
some clarifying text?