Subject: Re: erroneous ack packet, ideas please?
To: None <tech-net@NetBSD.ORG>
From: der Mouse <mouse@Rodents.Montreal.QC.CA>
List: tech-net
Date: 07/17/1997 23:29:33
> [[  eager accept ]]

> How does this compare to prior art, like the socket interface that
> Pink and Partridge used for their ST-2 implementation, which (IIRC)
> provides a vaguely OSI-like model where apps got to decide whether or
> not to accept a connection?

I'm not familiar with that, but I have used some vaguely prior art.  On
a certain old version of Ultrix, DECnet sockets (optionally) did more
or less what's being discussed here.  The interface, IIRC (it _was_
long ago), was that after creating a socket, you did a setsockopt() on
it before doing the listen.  When you did an accept, you got a socket
in a pre-accept state, and you had to setsockopt() it to either accept
it and make the socket usable or reject it and close the socket.

					der Mouse

		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B