Re: RF flag

Steve Deering <> Wed, 06 December 1989 00:08 UTC

Received: from by (5.54.5/4.7.34) id AA23976; Tue, 5 Dec 89 16:08:52 PST
Received: by; id AA21813; Tue, 5 Dec 89 16:08:45 -0800
Received: by Pescadero.Stanford.EDU (5.59/25-eef) id AA01576; Tue, 5 Dec 89 16:07:47 PDT
Date: 5 Dec 1989 15:44-PST
From: Steve Deering <>
Subject: Re: RF flag
Cc: "mtudwg" <mtudwg>
Message-Id: <89/12/05>
In-Reply-To: art's message of 5 Dec 89 152200 PST

> I think that Van might even agree that the DEC Bit has some advantages.

Actually, I don't think he would.  I've argued with him several times
about this issue, basically taking the same position as you, but I've
never heard him say a kind word about the DEC bit (except when compared
to doing *nothing* for congestion control).  I suppose we could ask him.

> The DEC Bit tends to detect congestion when it first appears and attempts
> to signal this to the end systems before the congestion gets out of hand.

You can do the same thing with packet loss, by intentionally discarding
some packets before it's actually necessary to do so.  I believe this is
what Van advocates.

Anyway, this is off the topic of this mailing list and out of my depth.
I just wanted to make the point that the DEC bit approach is controversial,
and I don't think it should have the only claim on the spare IP header flag.