BJ> Interesting....  I was wondering if Maximus's built in 
 BJ> ANSI (superset of VT-100 video) escape sequences would 
 BJ> cause any problems....

 BS> Maybe.. But after i set it to disable checking of ANSI, 
 BS> there might be no problem by my self. But i see i got a 
 BS> german caller thru, but i can see he didn't made it 
 BS> longer than the joining of the BBS.

Ok....

 BS> I notice, if calledbefore is set, there is no problem..

Ah.....  maximus has I believe two mecca files that are shown to a user on
the first few calls.  There is a new user (initial login sequence) and then
a thanks for comming back / here's some more startup info sequence.... 
Maybe Max is getting hung up on processing this second set of files?

 BJ> Are the TCP/IP ports set up for 8 bit clear or 
 BS> using 7 bit with various 
 BJ> terminal control codes interpreted?

 BS> I guess it's not 8bit, because things like filetransfer 
 BS> doesn't work.. but i don't know mutch about the telnet module.

Ok.  File transfers can get killed from either 7 bit or because of control
character interpretation.....  Concerning file transfers not working, the
standard telnet does not support maximus's file transfer protocols.  But if
you know what you are doing (or use a differnet terminal package) things
should work.....

 BJ> If the you run the 
 BJ> TCP/IP ports as the telnet standard with only 7 bits 
 BJ> and interpretation, then there are some VT-100 escape 
 BJ> sequences that *will* lock up the communication 
 BJ> path.  Gets back to the 
 BJ> old flow control issues, among other things.  

 BS> Well let me say it.. it look like a old school bbs in dos, after a printf
 BS> "\033(U".. I'm running ANSI, else i won't be able to 
 BS> write this message in MaxEd.

I figured some of that was hard coded.  Not nice in a unix enviornment
where a TERMCAP or CURSES library should be used instead (old school Unix
thoughts).  Now, with almost all folks emulating the terminal as a VT-100
work-alike, it is less of a problem......  Hmmm....  I wonder if the
exchange sequence includes a control-s from either end?  I recognize the
\033 as escape, and the (U is one of the functions (following an escape
character)....

 BJ> From 
 BJ> memory, control-s stops traffic while control-q starts 
 BJ> traffic.  Also (from memory), since Max is setup to 
 BJ> handle various modems, including flow control 
 BJ> issues, you may need to 
 BJ> look at the "modem" or "serial port" settings, or with 
 BJ> how you are openinging up the TCP/IP socket.  

 BS> Hmm.. ok. That's quite smart.. i used to run 
 BS> RemoteAccess, there i didn't got those options as far 
 BS> as i remember.

I've only run the BBS with hardware flow control due to the software flow
control issues.  But I have needed flow control with my setup using modems.

 BJ> For 
 BJ> TCP/IP connections we should be able to turn off the 
 BJ> software and probably also turn off the hardware 
 BS> flow control stuff in 
 BJ> Maximus.  For the TCP/IP socket, I believe you will 
 BJ> want an 8 bit clear channel, not the telnet interpreted 
 BJ> channel.  

 BS> I doubt that the telnet module is using the options, 
 BS> remember it's at alpha state.

Then you are probably defaulting parameters on opening up the TCP socket. 
I thought the default was 8 bit clean, and doesn't inturpet control
characters, so I'm probably barking up the wrong tree on this one.   [On
the other hand, if one option for opening the socket is to setup for
telnet, then 7-bit / flow control issues may be the problem.]

 BJ> There may be some additional IO CTL functions 
 BJ> also needing to be called after opening the TCP/IP 
 BJ> socket to get this 
 BJ> working smoothly.  [I assume we are running TCP and not 
 BJ> UDP over IP.]  

 BS> Correct.

 BJ> Is max being run as a shell under a version of getty?  

 BS> Maximus is listening (personally.. hah, couldn't find a 
 BS> better word) at a port number you tell it by the -p 
 BS> option, and after a user has login and he log out it 
 BS> quit, but then i've inittab to respawn the problem.

 BS> Maybe max should have some options, so i could be 
 BS> called from mgetty/inetd.conf, i really don't see the 
 BS> point of writing a serial IO interface, for modem 
 BS> connectivity... on unix..

I think eventually max needs to be able to take a hand-off from either
mgetty or inetd.  Max will need to be able to test (using the IOCTL
functions) for handling modem hardware flow control and to control DTR (to
hang up the phone) if it is to replace my existing setup.  The hooks should
already be in Max.  It's just a matter of figuring out (a) the IOCTL
functions to use and (b) wether they should be called....  Another item is
that Carrier detect from the modem should be tied to the TCP socket
disconnect for TCP based connections.

Take care.....

Bob Jones, 1:343/41

--- Maximus/2 3.01