Re: MRRU and MRU

sgw <sgw@sgw.xyplex.com> Fri, 03 February 1995 16:13 UTC

Resent-Date: Fri, 03 Feb 1995 10:53:23 -0500
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: sgw <sgw@sgw.xyplex.com>
Message-Id: <9502031550.AA13606@sgw.xyplex.com>
Subject: Re: MRRU and MRU
To: Tom Coradetti <tomc@digibd.com>
Date: Fri, 03 Feb 1995 10:50:09 -0500
Cc: gerry@spider.co.uk, ietf-ppp@merit.edu
In-Reply-To: <9502031534.AA02445@digibd.com>; from "Tom Coradetti" at Feb 3, 95 9:34 am
Resent-Message-ID: <"NQ0H_2.0.jQ1.m5bCl"@merit.edu>
Resent-From: ietf-ppp@merit.edu
Precedence: list
Resent-Sender: ietf-ppp-request@merit.edu

According to Tom Coradetti:
> 
> The theory was that the implementation of multilink as an LCP option forced
> the negotiation of SOME option, so that it might as will be implied by
> negotiating something useful. There was no ideal solution.
> 

Perhaps a better way would have been to have an LCP option called
"Multilink", with a 1- or 2-byte protocol field set to
"Multilink Protocol 1" == 0x1, followed immediately by the endnode identifier
(which can be null class for those who don't want to support what is currently
a separate option).

The idea of having a "protocol field" has precedence in how we implemented
LQM negotiation and VJ negotiation in LCP.

    +=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=+
    |        Scott G. Wasson              sgwasson@eng.xyplex.com |
    | Xyplex, Internetworking & Media     Voice:   (508) 952-4746 |
    | 295 Foster St. Littleton, MA 01460  Fax:     (508) 952-4887 |
    +=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=+

Re: MRRU and MRU sgw