Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT sources DOT redhat DOT com Delivered-To: mailing list cygwin AT sources DOT redhat DOT com Date: Mon, 16 Jul 2001 23:57:31 GMT From: Alison Winters Message-Id: <200107162357.f6GNvVU18358@sdf.lonestar.org> To: cygwin AT cygwin DOT com Subject: telnetd binary mode Hi, We have a piece of software that opens a telnet connection then initializes binary communication between server and client. It works happily on SCO OpenServer 5 and Red Hat Linux (5.2+). It doesn't on the latest Cygwin build. Essentially it sends a bit of information back and forth, then (seeming randomly) things hang. At first i thought it may be a telnetd version problem, but on SCO the daemon is built from 90/91/92 source files; the Cygwin build uses 93/95 source files. My suspicion is that this has something to do with binary vs text i/o. I know calling open(2) in Cygwin you need to explicitly add the O_BINARY flag; is something like this also the case when opening a socket? The other (related) possibility is that the telnet daemon should be set explicitly into binary mode (IAC WILL TRANSMIT-BINARY and IAC DO TRANSMIT-BINARY). This doesn't need to be done on UNIX, so i would find it odd if it had to be explicit in Cygwin, but is it possible that that is the problem? The final possibility is that there's a byte under 0x20 that is causing some kind of lockup. I've already escaped ^Q, ^S, ^], \r and \n to have it work correctly under UNIX; are there any that might be doing strange things in Cygwin specifically? I have set TERM to vt100 and done a tset before initializing the binary flow. Any help would be great appreciated. Thanks, Alison -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Bug reporting: http://cygwin.com/bugs.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/