From owner-freebsd-hackers Tue Dec 19 03:05:54 1995 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id DAA08125 for hackers-outgoing; Tue, 19 Dec 1995 03:05:54 -0800 (PST) Received: from hemi.com (hemi.com [204.132.158.10]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id DAA08117 for ; Tue, 19 Dec 1995 03:05:46 -0800 (PST) Received: (from mbarkah@localhost) by hemi.com (8.6.11/8.6.9) id EAA02661 for hackers@freebsd.org; Tue, 19 Dec 1995 04:05:40 -0700 From: Ade Barkah Message-Id: <199512191105.EAA02661@hemi.com> Subject: Help with telnet, flow control, and IXON. To: hackers@freebsd.org Date: Tue, 19 Dec 1995 04:05:40 -0700 (MST) X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-hackers@freebsd.org Precedence: bulk Hello, [Background] We have users logging into a FreeBSD 2.0.5-R machine by modem via a terminal server. Users connect from the terminal server to the FreeBSD box via `telnet', with the escape character supressed so they can transfer files (this all works fine.) [Problem] Users running certain applications on the FreeBSD box sometimes get garbled output. To be exact, the garbled output occurs soon after an application turns off IXON in the terminal's c_iflag. The user gets a string of garbled output, then everything fixes itself. Note, if a user telnets to the FreeBSD box via an intermediate site, the garbling disappears. It also never garbles when the user is running under `script'. [Question] My temporary fix is to recompile programs like telnet, etc., making sure IXON is left alone. I'm wondering what the "real" solution is. Another thing I can't resolve in my mind: the above shows that there's software flowcontrol going on between the terminal server and the FreeBSD box. But then, why do programs like `sz' work ? Is sz smart enough to escape and switch the START and STOP flow control characters ? And if I do let applications disable IXON, what other method of flow control can be put in place ? Hardware signaling is obviously out of the question over tcp/ip. I wonder why tcp/ip's own flow control algorithm doesn't work here. Thanks in advance for any hints... it's 4am and I'm pretty confused. =-) -Ade -------------------------------------------------------------------- Inet: mbarkah@hemi.com - HEMISPHERE ONLINE - www: --------------------------------------------------------------------