| |
PA2AGA > TCPDIG 31.08.96 04:03l 169 Lines 6747 Bytes #-10902 (0) @ EU
BID : TCP_96_175D
Read: GUEST
Subj: TCP-Group Digest 96/175D
Path: DB0AAB<DB0KCP<DB0ZKA<DB0LX<DB0RBS<DB0SEL<DB0ZDF<DB0SRS<DB0FP<DB0ERF<
DB0SHG<DB0HSK<DB0SGL<DB0HAG<DB0END<DB0ACC<PI8DRS<PI8DAZ<PI8GCB<PI8WFL<
PI8VNW
Sent: 960830/2236Z @:PI8VNW.#ZH2.NLD.EU #:58553 [Hoek v Holland] FBB5.15c
From: PA2AGA@PI8VNW.#ZH2.NLD.EU
To : TCPDIG@EU
Received: from pa2aga by pi1hvh with SMTP
id AA16545 ; Fri, 30 Aug 96 21:57:23 UTC
Received: from pa2aga by pa2aga (NET/Mac 2.3.61/7.1) with SMTP
id AA00003728 ; Fri, 30 Aug 96 22:50:20 MET
Received: from pa2aga-1 by pa2aga with SMTP
id AA00003690 ; Fri, 30 Aug 96 22:44:50 MET
Received: from pa2aga-1 by pa2aga-1 (NET/Mac 2.3.61/7.5.4) with SMTP
id AA00008006 ; Fri, 30 Aug 96 22:44:48 MET
Date: Fri, 30 Aug 96 22:38:39 MET
Message-Id: <tcp_96_175D>
From: pa2aga
To: tcp_broadcast@pa2aga-1
Subject: TCP-Group Digest 96/175D
X-BBS-Msg-Type: B
successful in creating a modem / system that uses this same setup. The user
can simply connect through the radio modem using PPP and access whatever
internet service is desired. The system also allows for a "light-weight"
protocol that allows for lower overhead for those services that do not need
it (actually, tcp is an example) that is more akin to what we know as KISS
mode communications. But, for what its worth, the system works, and it
works quite well.
One other point. The company is not interested in doing plug-in cards for
various computers. the one thing that is standard among computers is the
serial port. So, they use it at data rates as high as 128Kb. They also are
aiming at the portable computer market. Their service really does allow you
to use your laptop computer wherever you want (such as the long checkout
line at Fryes). Isn't this the vision we had oh so many years ago?
Seems we dropped the ball a few years ago by not persuing some of these
interesting ideas...
Jack Brindle
ham radio: wa4fib/6
internet: jackb@apple.com
------------------------------
Date: Wed, 28 Aug 96 10:08:03 GMT
From: "William Allen Simpson" <wsimpson@greendragon.com>
Subject: PPP/SLIP in connected AX25?
> From: alan@lxorguk.ukuu.org.uk (Alan Cox)
> > I strongly disagree with this view. PPP adds more data, but it isn't
> > "overhead" unless it does no useful work. With PPP, you get more
automatic
>
> The CRC's do no useful work because the HDLC in AX.25 does it.
>
PPP doesn't have CRCs. PPP itself (RFC-1661) is only:
1-2 bytes Protocol, variable Information
PPP is commonly run in HDLC (RFC-1662) framing, but is also run in X.25
framing (RFC-1598), or Frame Relay (RFC-1973). No reason why you
couldn't run it in AX.25.
I suspect that what you are thinking of is trying to run PPP as a "dumb
terminal" serial stream, HDLC framing inside AX.25 framing. I wouldn't
recommend it. Better to send on packet boundaries (someone has already
mentioned this).
> > configuration and considerably improved ease of use. True, we will likely
> > never need to exercise some of the enhancements of PPP, such as passing
IPX in
> > addition to IP, but we might be able to make use of them someday if we
avoid
> > closing off our options early.
>
One of those enhancements is data compression. When the application
isn't already compressing, this could be a big win here.
> Most PPP stacks use 15 or 30 second negotiation timers, you'll never
> negotiate over 1200 baud AX.25
>
The default PPP negotiation timer is 10 seconds. I've run it at 300 bps.
What problem do you think will occur at 1200 bps? The typical PPP LCP
packet is 10 bytes. And the timing should be configurable in any case.
WSimpson@UMich.edu
Key fingerprint = 17 40 5E 67 15 6F 31 26 DD 0D B9 9B 6A 15 2C 32
BSimpson@MorningStar.com
Key fingerprint = 2E 07 23 03 C5 62 70 D3 59 B1 4F 5E 1D C2 C1 A2
------------------------------
Date: Wed, 28 Aug 1996 14:30:01 -0700 (PDT)
From: Phil Karn <karn@qualcomm.com>
Subject: PPP/SLIP in connected AX25?
>I have serious reservations about the cable companies. I don't know about
>yours, but I have never yet seen a cable company for whom making a customer
>appointment was not a major upheaval in their day. If these outfits ever
hope
>to compete with telephone companies, they have a lot to learn about service.
I've had my doubts too. Recently an equally skeptical colleague paid a
visit to the local company, the one that's been digging up the streets
all over town. He came back impressed with what he saw. Redundant
fiber rings, natural gas fired backup generators in the equipment
huts, etc. The fiber-to-coax conversion happens at the 5-10 home
level. They finally seem to be getting serious about reliability and
signal quality.
>All cable modem ideas seem to depend upon technology which is grossly
unsuited
>to network access. In particular, they are highly asymmetric with lots of
>bandwidth to you and little bandwidth from you, and what bandwidth there is
to
>you must be shared. I am convinced that they are really secretly building an
>infrastructure for video on demand or -- worse -- home shopping.
Well, maybe. For some time, the joke was that the cable companies
defined "two way cable" as placing a "BUY" button on the TV remote
control. But that was before they discovered the Internet.
I've seen a variety of data rates quoted. They are indeed shared and
asymetrical with less capacity in the reverse direction, but even the
slowest figure I've seen (768kb/s) is quite a bit faster than an ISDN
channel. The downstream rates range anywhere from 6-27 megabits/sec
per TV channel allocated to the system. And I don't really mind the
asymmetry because even though I leave my machine up all the time I
don't really want to turn it into a web server. Like just about
everybody else, I receive a lot more than I send.
Re contention sharing, this is not exactly a new concept -- Ethernet
has done it quite successfully for years. The local company says
they're engineering 250-500 homes per subnet. That's homes passed by
cable, not necessarily homes using the cable modem
service. Extrapolating from the number of homes near me with ISDN (and
from Qualcomm's hiring rate), I expect it would still be a while
before there would be as many as 10 other customers sharing the subnet
in my development. And they probably wouldn't all be actively
downloading the Linux source distribution at the same time.
Phil
------------------------------
Date: 28 Aug 96 12:46:20 -0700
From: "Jack Brindle" <jackb@apple.com>
Subject: PPP/SLIP in connected AX25?
--Cyberdog-AltBoundary-0006D067
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
>The local cable company (which has been trenching the streets just
>about everywhere for the past year) is mumbling something about cable
>Internet access coming to San Diego by the end of the year. They're
>tossing around prices like $19.95 or $29.95/mo flat rate, modem
>included.
>
Wow. Add that to my monthly $33 bill from TCI, and I'd be sending $60
To be continued in digest: tcp_96_175E
Read previous mail | Read next mail
| |
