Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!nstn.ns.ca!news.cs.indiana.edu!mips!swrinde!zaphod.mps.ohio-state.edu!usc!ucla-cs!ucla-se!PRICE@uclapp.physics.ucla.edu
From: price@uclapp.physics.ucla.edu (John Price)
Newsgroups: comp.sys.mac.comm
Subject: ZTerm transfers (again...)
Message-ID: <009496DA.C7E09C00@uclapp.physics.ucla.edu>
Date: 31 May 91 21:18:15 GMT
Sender: news@SEAS.UCLA.EDU
Reply-To: price@uclapp.physics.ucla.edu (John Price)
Organization: UCLA Particle Physics Research Group
Lines: 44

Hi all:

I have a question about just how ZTerm does file transfers.  I transfer 
files from a VAX running VMS (rz 3.01, sz 3.03) to either a Plus (at 2400 
bps) or a IIcx (at 19200 bps) and back.  Transfers *from* the VAX work 
great - I get 237 cps (!) to the Plus, and about 1050 cps to the IIcx.  I 
don't expect to be able to improve either speed by too much - certainly not 
the Plus...

However, going from either Mac to the VAX is painful.  I keep getting ZRPOS 
errors.  The fastest transfer I can get on the IIcx is about 500 cps.  I 
tried playing with the Zmodem parameters a bit - I noticed that changing 
the "subpacket length" didn't have much effect, but changing the "Window 
limit" did seem to help.  

Initially, there was no window limit - the box was empty.  Not having the 
slightest idea what I was doing, I put 100 there and tried the transfer 
again.  This time, it appeared as if the program wasn't trying to send as 
much data at one time as it had before - the "bytes sent" field updated 
much more quickly than before.  When I looked at the window limit again, to 
try a different number, I saw that the program had automatically changed it 
to 256.

OK, I can deal with that.  It wants multiples of 256.  I have no idea why, 
but now I at least have a handle on how to affect the program's operation.  
I found that the larger I made this number, the faster I was able to 
transfer data.  This effect seems to flatten out with a window limit of 
about 4096 (it's currently set to 16384).

Now, my questions to Dave Alverson (or anyone who has more of a clue about 
this than I do - that shouldn't cut too many people out  :) are the 
following:

	1: What is the "subpacket length"?
	2: What is the "window limit"?
	3: How can I optimize the transfer speed *to* the VAX without 
	   affecting the transfer speed *from* the VAX?
	4: (to Dave Alverson specifically) Could you put at least a cursory 
	   description of this in the docs for the next version?  

Thanks a lot in advance!

           John Price * * * * price@uclapp.physics.ucla.edu
           Where there is no solution, there is no problem.