Path: utzoo!attcan!uunet!cs.utexas.edu!tut.cis.ohio-state.edu!ucbvax!UCONNVM.BITNET!SEWALL
From: SEWALL@UCONNVM.BITNET (Murph Sewall)
Newsgroups: comp.sys.apple
Subject: Re: File Transfer Problems
Message-ID: <8909280919.aa18481@SMOKE.BRL.MIL>
Date: 28 Sep 89 14:07:53 GMT
References: 
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: School of Business - U. of Connecticut
Lines: 36

On Tue, 26 Sep 89 22:39:28 EST you said:
>When x-fered from an Apple to a Mac, the mac registers MORE blocks coming
>through than the Apple registers going through.  It does NOT affect the x-fer
>in any way that I have found as the text file was sent error free.

If you're using Kermit-65 on the Apple, it counts blocks in HEX.  I'm not
sure about the Mac side, but it wouldn't be the first Kermit implementation
I've seen that counts blocks in decimal.

>I thinkI remember hearing that Kermit was made specifically for different
>computer x-fers, but my sources may be wrong.  All I know is that it seems
>slower that a normal x-fer on the Apple side and FASTER on the Mac side.

The original problem was: most mainframes use either mark or even parity,
leaving only 7 bits for data (conceived of LONG before anyone thought that
commputers would 'talk' to anything other than dumb-terminals and other ASCII
devices), but users need to transfer 8-bit files.  Kermit was designed to
be able to encode 8-bit data into 7-bit characters AND transfer over packet
oriented networks.

The speed HAS to be the same on both ends (any differences are illusions),
but the Kermit protocol has all sorts of optional features (which have to
be implemented on both ends to work, of course) - 3 levels of CRC error
checks, long packets (up to 1K?), 'sliding windows', etc.  So, depending
on what's transfering to what with what features active, the wall clock time
for a transfer can vary somewhat :-)

Murph Sewall                       Vaporware? ---> [Gary Larson returns 1/1/90]
Prof. of Marketing     Sewall@UConnVM.BITNET
Business School        sewall%uconnvm.bitnet@cunyvm.cuny.edu         [INTERNET]
U of Connecticut       {psuvax1 or mcvax }!UCONNVM.BITNET!SEWALL     [UUCP]
           (203) 486-5246 [FAX] (203) 486-2489 [PHONE] 41 49N 72 15W [ICBM]

    The opposite of artificial intelligence is genuine stupidity!
-+- I don't speak for my employer, though I frequently wish that I could
            (subject to change without notice; void where prohibited)