gallo@cs.albany.edu (Andrew Gallo) (05/17/91)
I'm a casual Win 3.0 user, and plan to buy a new 386 so I can run it. I use my PC mostly for terminal connections to my university's Unix machines. Wouldn't it be nice if there was a way for you to have multiple widows open on a remote machine, like X allows you to do but via modem. Imagine this: a client sits between your terminal program and your COMM port. You then open multiple instances of Kermit (or whatever term program) on your PC. Every time Kermit tries to write to the COMM port your client interupts and intercepts the information, decides where the data came from (each instance of Kermit is given an ID), and encapsulates a packet. The packet is sent over modem to a server running on the Unix machine which maps Window ID's to Unix process ID's (pid's). The server would be waiting on /dev/ttyxx for i/o, unpack the packet, determine for which process the data is intended, and send it down a pipe to that process. Likewise processes would write to a pipe to the server who would encapsulate a packet destined for a particular window on the PC side. This *should* allow a Windows user to visually and interactively send/receive i/o from multiple Unix processes, simulating the X functionality. There are a couple of problems: bootstrapping (switching from character transport to packet transport after an initial connect is made), mapping pid's to Window ID's, catching Kermit's attempt to write to a COMM port (although I'm sure an interupt handler could be written to do this). So I'm asking for your educated opinion - is this doable? Has someone beat me to it? If so, who, and where can I get the program? If not, any suggestions on how to implement this design? Thanks for input. If this is not discussion that should be posted on the net, feel free to send me private email at gallo@cs.albany.edu. >-----------------------------------------------------------------------------< > Andy Gallo | State University of | "If six, turned out to be < > gallo@cs.albany.edu | New York at Albany | nine, I don't mind..." 8-) < >-----------------------------------------------------------------------------<
leoh@hardy.hdw.csd.harris.com (Leo Hinds) (05/17/91)
In article <834@hartmanis.albany.edu> gallo@cs.albany.edu (Andrew Gallo) writes: > Wouldn't it be nice if there was a way for you to have multiple >widows open on a remote machine, like X allows you to do but via modem. Recently (on comp.binaries.ibm.pc ?) the "infamous" UW - PC port was posted. The features are basically what you describe, but under the "native DOS" environment. Whether or not this will run as is under Win3, or more ideally the author will port it to Win3 are two questions I do not have answers for ... anybody out there know? leoh@hdw.csd.harris.com Leo Hinds (305)973-5229 Gfx ... gfx ... :-) whfg orpnhfr V "ebg"grq zl fvtangher svyr lbh guvax V nz n creireg ?!!!!!!? ... znlor arkg gvzr
otto@tukki.jyu.fi (Otto J. Makela) (05/17/91)
In article <834@hartmanis.albany.edu> gallo@cs.albany.edu (Andrew Gallo) writes:
[are there multiwindow Unix terminal programs for Windows3 ?]
I haven't seen a Windows-based version of UnixWindows (which is originally a
Mac-to-Unix protocol), but there is a text-screen based implementation for PCs.
Look at file pd:<msdos.modem>uwpc105.zip on wsmr-simtel20.army.mil
Would some Windows guru like to implement it for Windows ?
--
/* * * Otto J. Makela <otto@jyu.fi> * * * * * * * * * * * * * * * * * * */
/* Phone: +358 41 613 847, BBS: +358 41 211 562 (USR HST/V.32, 24h/d) */
/* Mail: Kauppakatu 1 B 18, SF-40100 Jyvaskyla, Finland, EUROPE */
/* * * Computers Rule 01001111 01001011 * * * * * * * * * * * * * * * * */bchen@wpi.WPI.EDU (Bi Chen) (05/20/91)
I wounder if a 2400 bps modem would ever be fast enough to make a 386 an X terminal? Isn't it impossible or is it?
dsims@uceng.UC.EDU (david l sims) (05/20/91)
bchen@wpi.WPI.EDU (Bi Chen) writes: >I wounder if a 2400 bps modem would ever be fast enough to make a 386 >an X terminal? Isn't it impossible or is it? I recall reading that X Windows requires at least a 19.2K serial line to run at a "just okay" level. Now since 2400 is approximately eight times slower than 19.2K, you make the decision. It would be interesting to see X in action on a 2400 line....
splin@nima.berkeley.edu (Steven Lin) (05/21/91)
In article <8524@uceng.UC.EDU> dsims@uceng.UC.EDU (david l sims) writes: >bchen@wpi.WPI.EDU (Bi Chen) writes: > >>I wounder if a 2400 bps modem would ever be fast enough to make a 386 >>an X terminal? Isn't it impossible or is it? > >I recall reading that X Windows requires at least a 19.2K serial line to run >at a "just okay" level. Last summer when I was dabbling with SLIP, I decided to see how well X would run over a serial line. I connected an X Terminal to a Sparcstation 1+ through the serial port running at 9600 baud. The performance was barely adequate. Opening and moving windows were intolerably slow. The terminal would virtually freeze if it ever required a font that wasn't present. I can't imagine what life would be like at 2400 baud.
chris@visionware.co.uk (Chris Davies) (05/21/91)
>bchen@wpi.WPI.EDU (Bi Chen) writes: >I wounder if a 2400 bps modem would ever be fast enough to make a 386 >an X terminal? Isn't it impossible or is it? In article <8524@uceng.UC.EDU> dsims@uceng.UC.EDU (david l sims) writes: >I recall reading that X Windows requires at least a 19.2K serial line to run >at a "just okay" level. Now since 2400 is approximately eight times slower than >19.2K, you make the decision. It would be interesting to see X in action >on a 2400 line.... No it wouldn't. It's about as interesting as watching paint dry. Chris -- VISIONWARE LTD, 57 Cardigan Lane, LEEDS LS4 2LE, England Tel +44 532 788858. Fax +44 532 304676. Email chris@visionware.co.uk -------------- "VisionWare: The home of DOS/UNIX/X integration" -------------
arasmith@mathcs.emory.edu (David Arasmith) (05/22/91)
In response to questions regarding UWPC, it seems to run fine as a DOS
task in Windows (386 enh mode). I got this from c.b.i.p a while ago and
asked the author if he was considering a port to Windows. His response wato say he was waiting on Borland C++ and then Windows would be a happenin thang.
You might want to check with him (I don't remember his name right now -
although I remember he is in Australia).
Check UWPC out. It's pretty convenient (I think it even includes the
host for the UNIX side).
--
David M. Arasmith | arasmith@mathcs.emory.edu Internet
Emory University | {sun!sunatl,gatech}!emory!arasmith UUCP
Dept of Math and CS |
Atlanta, GA 30322 | I should be working! Gee....I wonder what's on TV?dmoffatt@silver.ucs.indiana.edu (David Moffatt) (05/27/91)
In also had the same idea. I am a computer science student at Indiana Universtiy and one of the things we have to do is senior projects so that is what doing. So far the program look nothing like what you described. I doubt I will ever get anything that would be a useable also. But as far as I can see so far it is quit a simple matter to time slice the serial line useing small packets and giveing priority to keyboard input. The way it works in my scheme you can download files in the background but that is about it. So far the biggest snafu I have had to deal with are the user commands like remote dir. This is not a hard problem it just requires a little more work than putting the files into packets. I think when someone comes out with a good program to do this it will quickly take the market by storm. But the program has to easly ported to mainframes and has to use a good scheme of prioritizing. Also it will be incompatable with present day stuff or maybe it could be made so that it worked like a normal com. program until a special command was gives to spawn additional windows? Oh well ... who knows what may happen. Dave Moffatt the hack who is not a geek.