home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.archives.msdos.d
- Path: sparky!uunet!psinntp!panix!carlf
- From: carlf@panix.com (Carl Fink)
- Subject: Re: *** WHAT'S THE BEST COMM PROTOCOL ***
- Message-ID: <1992Dec23.015825.16809@panix.com>
- Date: Wed, 23 Dec 1992 01:58:25 GMT
- References: <1992Dec21.181829.4391@porthos.cc.bellcore.com> <724973743snz@chrism.demon.co.uk>
- Organization: PANIX Public Access Unix, NYC
- Lines: 38
-
- In <724973743snz@chrism.demon.co.uk> chris@chrism.demon.co.uk (Chris Marriott) writes:
-
- [deletion]
-
- >It is easy to say what you should NOT be using, and that's Kermit
- >or XMODEM. These two protocols have a VERY high overhead, and were
- >really designed for slow (300 baud or so) comms over noisy lines.
-
- >As to what you SHOULD be using - that's tougher, and depends on your
- >modem. If you have a modem which *doesn't* do error correction
- >itself, you should use an error-correcting protocol with a long packet
- >length - YMODEM or XMODEM/1K are the most common. If, as most modern
- >modems do, your modem itself does error detection/correction (ie, supports
- >MNP4 or V42), you'd be better off with a protocol which assumes an
- >error-free line. Of these the most common is ZMODEM, although slightly
- >more efficient, if you have software which supports it, is YMODEM-G.
-
- >I hope this has not added to your confusion!
-
- >Chris
- [sig deleted]
-
- ZMODEM is faster than YMODEM in every circumstance I've tried them
- in, and also has better error correction and other advantages. Why
- do you recommend YMODEM for non-error-correcting modems?
-
- The following not directed at Chris: someone commented that ZMODEM
- is slowed down by using too small a block size. Of course, ZMODEM
- is a streaming protocol, and doesn't use blocks. You can vary the size
- of the units being CRC checked, but according to Chuck Forsberg (who
- invented the ZMODEM-90 protocol) it won't affect transfer efficiency by
- more than half of one percent.
-
- --
- Carl Fink carlf@panix.com, C.FINK4(GEnie), or CF427620I@LIUVAX.BITNET
- "Facts are stubborn things; and whatever may be our wishes, our
- inclinations, or the dictates of our passions, they cannot alter
- the state of facts and evidence" -- John Adams
-