Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!linus!philabs!seismo!hao!hplabs!sri-unix!cc.fdc@COLUMBIA-20.ARPA From: cc.fdc@COLUMBIA-20.ARPA@sri-unix.UUCP Newsgroups: net.micro.cpm Subject: none Message-ID: <4691@sri-arpa.UUCP> Date: Tue, 30-Aug-83 09:46:15 EDT Article-I.D.: sri-arpa.4691 Posted: Tue Aug 30 09:46:15 1983 Date-Received: Fri, 2-Sep-83 17:41:33 EDT Lines: 12 From: Frank da Cruz We haven't done thorough studies, but we find that in a micro-to-mainframe setting, in which KERMIT is usually used, the efficiency varies between 50 to 80 percent (user bits/baud rate). I don't know about CIS protocol A (what is it?), but KERMIT has somewhat shorter blocks than MODEM and does various kinds of quoting and prefixing of special characters to accommodate mainframes that can't take in binary (8-bit) data or even certain ASCII control characters. The actual amount of overhead varies with the capabilities of the two hosts (the two KERMITs will configure themselves to each other) and the speed of the slower host (timesharing systems can get very slow...). - Frank -------