Path: utzoo!utgpu!watserv1!maytag!xenitec!zswamp!root
From: root@zswamp.fidonet.org (Geoffrey Welsh)
Newsgroups: comp.dcom.modems
Subject: Re: Overseas Connections with high-speed modems
Message-ID: <6929.27D32625@zswamp.fidonet.org>
Date: Mon, 04 Mar 91 10:55:27 EST
Organization: Izot's Swamp BBS - Kitchener, Ontario

Pat Myrto (pat@rwing.UUCP ) wrote:

 >Oh, yes there ARE protocols to "spoof" in the MSDOS world - 
 >virtually
 >all the non-streaming (ACK-NAK, or 'send and wait') file 
 >transfer
 >protocols take a big performance hit with the hi-speed 
 >modems, even
 >with non-PEP modem protocols, probably because of the 
 >packet-type
 >nature of the error-correction schemes, such as MNP.

   You are correct that any non-streaming protocol will suffer at high speeds, 
but packetizing via MNP or X.25 simply make the problem worse, not cause it.

   Think of XMODEM: send a block, get the ACK, send the next block.  At 300 
baud, the 128-byte block takes several seconds to transmit, the pause while 
waiting for the ACK only a fraction of a second, therefore the fraction of 
transmitter time which remains unused is small.  Now go to 38,400 bps, where 
the pause for the ACK is longer than the time it takes to send a block... the 
result is that XMODEM will usually remain below 1000 CPS, *no matter how high 
the baud rate goes*

   You can use larger blocks, getting more data through for a fixed wasted 
time, but that only reduces the overhead as a percent of the available 
transmit time and does not solve the problem.

   You can use windowed protocols, but you must impose an arbitrary limit on 
the size of the window and, when the baud rate reaches a certain point, a full 
window will be transmitted before the first ACK comes back, making the 
protocol functionally equivalent to XMODEM, but with an effective block size 
equal to the maximum window size.

   The long term solution is the streaming protocol, and spoofing is an 
effective conversion of a non-streaming protocol to a streaming protocol for 
the physical transport and then back again at the other side.
 

--  
UUCP:     watmath!xenitec!zswamp!root | 602-66 Mooregate Crescent
Internet: root@zswamp.fidonet.org     | Kitchener, Ontario
FidoNet:  SYSOP, 1:221/171            | N2M 5E6 CANADA
Data:     (519) 742-8939              | (519) 741-9553
The mile is traversed not by a single leap, but by a procession of coherent 
steps; those who insist on making the trip in a single element will be failing 
long after you and I have discovered new worlds.        - me