[comp.sys.apollo] VT100 Emulation/crp problem.

trb@stag.UUCP ( Todd Burkey ) (12/10/87)

Hi,
  I have been having a problem with trying to run programs like vi
on other systems from my Apollo workstation. Basically, what I recently
tried doing was:

  1) vt100                 <-enter the vt100 emulator.
  2) crp -on e4c0 -me      <-move over to a node that has telnet running
  3) telnet bozo           <-connect to another node (in this case a Sun)
  4) login to bozo
  5) tset vt100            <-set terminal type to vt100
  6) set term=vt100        <-   ""
  7) vi test.c

Now, the screen addressing seems to work ok and the file loads up ok, but
I am still in a mode where I have to enter a return after each command I
want to execute (i.e. the j,k,l,h, etc keys...) There must be some way of
getting the vt100 emulator to transmit a character at a time, but I am at
a loss as to how to do this! It isn't anything related to the system I log
into, but rather the emulator (or telnet), buffering by line. Has  anyone
else had this problem and solved it? Also, the Apollo appears to be sending
a cr/lf pair (which screws up other screen programs...)

   -Todd Burkey
    trb@stag.UUCP

joey@tessi.UUCP (Joe Pruett) (12/15/87)

The problem is crp.  It only supports a line at a time mode.  Complain
to apollo, and it might get fixed some year.  One other possibility
is to use the TCP/IP package and telnet or rlogin.  Both of those
programs are character at a time interfaces.

	-joey

holtz@cascade.carleton.CDN (Neal Holtz) (12/17/87)

> Sender:     Joe Pruett <tektronix!reed!tessi!joey@ucbvax.berkeley.edu>

> The problem is crp.  It only supports a line at a time mode.

Not quite.  As with a lot of Apollo stuff, the problem isn't with
missing functionality, it is with missing documentation (not a whole
lot better, I agree).

Programs running in a CRP'ed process can configuring the pad on the
controlling node to be in raw mode, can program the function keys, etc.
One can look at the source for EMT to see how this is done.

For example, I hacked an early version of C-Kermit so it runs properly
(i.e., gives the standard C-Kermit (and TOPS-20) character-at-a-time
interface) from an sio login, from a pad, from a CRP'ed process and from
under the vt100 emulator.  Unfortunately, I can't seem to get it to
work when CRP'ed from an sio login nor when CRP'ed from a VT100 emulator.
In case any one is interested, I have about 370 lines and 14K of C code
I can send you, showing how its done.

Apologies: I lost the thread on what the original article mentioned,
so this may not answer any of the original questions.

----------------------------------------------------------------------------
Prof. Neal Holtz                    (Until Internet registration completed:)
Dept. of Civil Engineering     CDNnet:   Holtz@cascade.carleton.cdn  
Carleton University            CSNET:    Holtz%cascade.carleton.cdn@ubc.csnet
Ottawa, Ontario         ARPA?: Holtz%cascade.carleton.cdn%ubc.csnet@relay.cs.net
Canada  K1S 5B6
(613) 564-2677, -4315 (msgs)