TIP: Click on subject to list as thread! ANSI
echo: cis.os9.68000.osk
to: Bob van der Poel 76510,2203 (X)
from: Mark Griffith 76070,41
date: 1992-11-08 21:04:52
subject: #16949-CIS and 9600 bps

#: 16988 S12/OS9/68000 (OSK)
    08-Nov-92  21:04:52
Sb: #16949-CIS and 9600 bps
Fm: Mark Griffith 76070,41
To: Bob van der Poel 76510,2203 (X)

Bob,

 > BTW, is 1k the recommended buffer size for B+ at 9600? Or would
 > something a bit larger be better? Does sterm reduce the size
 > dynamically if errors start to get out of hand (or does the protocol
 > even support that kind of thing)?

B+ will back down to a "send and wait" protocol (like old B or XMODEM) if
enough errors occur, and then go back to full B+ when the transmission starts
going well enough again.

B+ also specifies that the packet size is 1024 bytes at 2400 baud or above, 512
bytes for 1200 and 128 for 300 baud.  Their reasoning is to keep the packet
send time at 5 seconds or less to help reduce the cost of sending repeat
packets in case of an error, and also to reduce the amount of time it takes for
the protocol to recover from on error when there are several packets in the
acknowledgement queue (since B+ is a sliding windows type protocol).

BTW: Steve Wegert and I have the OSK version of Kermit 5A which also has
sliding windows.  Works real nice.  Now if CIS or Delphi would implement it it
would be better.

Mark

SOURCE: compuserve via textfiles.com

Email questions or comments to sysop@ipingthereforeiam.com
All parts of this website painstakingly hand-crafted in the U.S.A.!
IPTIA BBS/MUD/Terminal/Game Server List, © 2025 IPTIA Consulting™.