home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!uunet.ca!yuggoth!rbharding
- From: rbharding@yuggoth.org (Ron Harding)
- Newsgroups: comp.sys.amiga.datacomm
- Subject: Re: Modem Terminology is Confusing
- Distribution: world
- Message-ID: <rbharding.038j@yuggoth.org>
- X-NewsSoftware: GRn 1.16e (7/4/92) by Mike Schwartz & Michael B. Smith
- Date: 22 Dec 92 20:52:40 EST
- Organization: The Leningrad Carrot Conspiracy
- Lines: 101
-
- In article <Bzo8y3.M4H@news.cso.uiuc.edu> ja51359@uxa.cso.uiuc.edu (axelrod) writes:
- > Could someone please explain in a followup message what the various protocols
- > mean? i.e. what is v.42/v.32, what do the various MNP levels mean? Is a
- > 14.4K bps modem really 14.4K on purely random data? What are the capabilities
- > of each protocol when sending data both directions at the same time?
- >
- > Thanks,
- > Jeff
-
- Snicker. Very confusing indeed. Somebody has fucked up royally with the
- names. Anyway, here's the info:
-
-
- V.32 : method for transmitting raw data at 9600 bps full-duplex (both
- directions at once.)
-
- V.32bis : method for transmitting raw data at 14400 bps full-duplex.
-
- HST 9600 : method for transmitting raw data at 9600 bps in one direction,
- and 2400 bps in the other direction. A short delay is required
- to the turn the line around. This is usually not a problem
- in most terminal sessions or while downloading, since data flow
- is quite lopsided in either case. Unfortunately, it's
- proprietary to US Robotics, and only found in their modems.
- It's quite common now, but with the increasing popularity of
- V.32xxx modems, I think HST's days are numbered.
-
- HST 16800 : I don't know anything about this. Only heard of it a couple
- days ago. Probably still proprietary.
-
- V.42 : error-control protocol. Should simulate an error-free connection.
- Also known as LAPM (Link Access Procedure for Modems).
-
- V.42bis : data compression protocol. Achieves up to 4x compression, giving
- a max throughput of 57600 bps using V.32bis. Also known as BTLZ
- (British Telecom Lempel Ziv).
-
- MNP2-4 : Various error-control protocols. I don't know the differences
- between them. "Microcom Network Protocol"
-
- MNP5 : data compression protocol. Achieves up to 2x compression.
- Unfortunately it hurts your throughput if you try to send
- uncompressible data through it.
-
- In addition to that list, there are also fax protocols, about which I know
- very little.
-
- A V.32 modem can achieve 9600 bps even on random or compressed data. A 2400
- bps modem using V.42bis, even though you may set your terminal program to
- 9600 bps, can only achieve 9600 bps with very compressible data. Like a
- file of all zeroes.
-
- In many cases, you don't need to put all your eggs in one protocol basket.
- Hayes Optima 144 supports everything up there except the HST standards, and
- optionally FAX protocols.. The US Robotics HST Dual Standard supports at
- least HST and the V.32 suite, and probably others but I don't know for
- certain.
-
- Since we're on the subject, I'll also talk a bit about modem escape sequences.
- Escape sequences are sent to the modem to change it from data-transmission
- mode to command mode, usually to issue an "ATH" (hang up) sequence.
-
- Hayes Improved Escape Sequence with Guard Time (what a mouthful):
- Consists of short delay, followed by three special characters, usually
- '+', followed by another short delay. Patented (and vigorously enforced)
- by Hayes.
-
-
- TIES : "Time-independant Escape Sequence" Used by Hayes-pseudo-compatible
- modems. The sequence consists of the five characters: "+++AT",
- that is, the Hayes sequence, but without the guard times, followed
- by the "AT" characters that must preceed any Hayes-compatible
- command sequence.
-
-
- The debate about escape sequences centers on the fact that there is no
- guarantee that the TIES sequence won't occur accidentally in the middle of
- sending some binary file. Anything can happen in the random-looking
- garbage produced by a compression program. When a random "+++AT" shows up
- in the data stream, the modem will pop into command mode, and try to interpret
- the rest of the data as modem-commands.
-
- The probability of this happening is usually very small. But consider this:
- were I using a TIES modem, I wouldn't be able to send out this posting. It
- contains two TIES escape sequences. I'm using a Hayes Optima 144, so even
- though this file contains the "+++" sequence (there I go again), it won't
- have the required delays before and after the '+'s (unless, of course, by
- some odd circumstance, my UUCP transfer gets preempted by a higher priority
- task before and after sending the "+++". Even the Hayes sequence isn't
- perfect.)
-
- In the spirit of complete honesty, let me confess here that I'm a Hayes
- employee, albeit in LAN software, not modems. I've written this on my own
- time, on my own system, in my own bedroom while drinking my own homebrew, so
- nothing I've said is on behalf of Hayes. Any errors are likely the fault of
- the homebrew.
-
- +---------------------------------------------+--------------------------------+
- | Nuke'Em: Get them before they get you! | Ron Harding |
- | Another quality home game from Butler Bros. | uunet.ca!yuggoth!rbharding |
- +---------------------------------------------+--------------------------------+
-