SSI/EIA for 630 et. al. (was: Re: AT&T 630 terminal)

Robert C. White Jr. rwhite at nusdhub.UUCP
Tue Jan 24 15:22:32 AEST 1989


in article <1762 at edison.GE.COM>, rja at edison.GE.COM (rja) says:
> 
> In article <1003 at vsi.COM>, friedl at vsi.COM (Stephen J. Friedl) writes:
> The AT&T terminals catalog in front of me does not refer to the
> SSI/EIA board and specifically indicates that 2 RS-232-C ports
> are included with the #3344-630 controller.  The 122-key keyboard
> is shown as retailing at $150 now.   All told it still looks like
> around $2500 retail from AT&T.

In normal operations the 630 is connected to a computer by the "main"
port leaving the "auxillary port" free for other purposes.  Since the
630 can support two concurrent hosts with up to 7 layers-windows each
(total of 14 host windows) and as many "local" windows as you can fit in
memory.  (A "local" window is created by (1) downloading an application
into the terminal memory and executing it and then having it close the
connection back to the computer (all supported via the jx protocol under
layers and the applications programmers toolkit for the 630); or
(2) running a "cached" program form the "more" menu (Programs are
chached by (a) JXing them and having them cache themselves, or having
them available via pull-out cartridge [like a video game, some few of
which for the 630 are available to persons at bell labs ;-)])

The SSI/EIA Board is a daughter board for the 630 which possesses two
physical connectors; the first connector is a female DB-25 connector and
the second is a 4-pair modular jack supporting the 56kbs SSI protocol
AT&T uses with it's SNA cluster controllers.  Only one of these ports may be
used at a time.

The purpose of this card is to allow you to use both the second host
capibility and a locally attached printer.  The printer support is pretty
snappy, as is the dual host capibility.  To use both you need the extra
port.

At the present time I am not aware of any SSI device for the 630 to talk
to.  The 6500 cluster controllers do not recognize the SSI
device-type-signature of the 630 (though our software is a bit out of
date on our 6500's I understand from my account rep that the newer
release is unchanged in this respect) and I am not aware of an SSI board
for any of their computers or their ISN.

Overall I would rate the 630 as fairly snappy with one major gripe:
If you press-and-hold a mouse button, all other activity on the terminal
stops.  No data is lost but the time is and if the buffer fills the
application will wait on flow control.  This major nusiance is also true
for the big-time windowing applications like new or reshape (if you
select a multi-step operation like reshape, work stops until you finish
or cancel).  The terminal definitely has the processing power for this
not to be a problem, so I strongly suspect that the problem is the way
these operations are handled is in the layers software.

The JX loader/support program (which runs on the host to download
applications and hangs around until dismissed by the program it loaded
into the 630) provides the hooks necessary to implement the unix file
operations so a program written for the 630 can access any files and
devices the user can normally.  Programs like EMACS, spreadsheets, or
whatever can be ported directly into the 630 environment, making the
device exteremely flexible.  The biggest drawback to the
630-as-diskless-workstation is the 19200 baud maximum communications
speed.  This may be what the SSI port is to be used for eventually,
meanwhile you just watch the application "fill up" the window while the
mouse icon pretends it's a coffee cup.

In playing the afore-mentioned video games on a 630 (during a u3g
conference demonstration) proved to me that the interrupt abilities and
screen refresh abilities are better than you might think just from using
layers.  With local windows running and host operations sitting in
backround I ran GBACA ( Get Back At Corporate America, a corporate
astroids game with cloning AT&T and Bell logos (as in break-up), Mtv
throwing notes, "Crazy Eddie" spitting advertisements, an IBM that goes
three separate ways when you shoot at it, and the occasional Golden
Arches food offering [There was supposedly more int there but I didn't
get the chance to get good enough at the came to see any more of it] all
being moved around the screen with amazing density and speed) and didn't
find the mouse interrupts (used for moving, firing, and so forth)
slowed the performance to any noticeable degree.

In general, the display update and processing speed beat the pants of
of a lot of graphic devices I have seen out there, but until something
better than layers comes along the terminal is over-powered compared to
the actually available interface and software.

Rob.



More information about the Comp.unix.wizards mailing list