Date: Wed, 24 Nov 1999 20:27:26 -0500 (EST) From: Brad Karp <karp@eecs.harvard.edu> To: freebsd-mobile@freebsd.org Subject: Re: STRIP (was Re: richochet modems) Message-ID: <199911250127.UAA11786@dominator.eecs.harvard.edu>
next in thread | raw e-mail | index | archive | help
[forgot to cc -mobile on my reply to Matt] To: matt@braithwaite.net Subject: Re: STRIP (was Re: richochet modems) > Right. STRIP does this too, but it's a kludge: a central node with a > known hardware address has to be the ``ARP server'' for everybody > else. Yes, my point exactly: HUMR requires no central ARP server. > Sorry, can you explain that a little more? Are you saying that given > any two radios, you can set up a reliable byte stream between them > using AT commands? Say you have two hosts, A and B, each with a Ricochet, MAC addresses 0000-0001, and 0000-0002, respectively. B can tell its radio: ATS0=1 A can tell its radio: ATDT0000-0002 This will *literally* emulate a plain-old-modem-style byte stream connection from host A to host B. You can just use standard client and server PPP on the hosts on either side, then. No STRIP required. So my overall point is: for a single hop of PPP over Metricom, there's no need to use STRIP at all. And if you want multi-access, or multi-hop, the central ARP server STRIP requires makes HUMR more convenient. About the newest radios' MAC addresses: yes, the string parsing routines need to be changed accordingly. I've not done this for HUMR yet, either. -Brad, karp@eecs.harvard.edu To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-mobile" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199911250127.UAA11786>