Computone grumbling

[Bill Vermillion]

In article <1232 at swusrgrp.UUCP> jeff at swusrgrp.UUCP (Jeff Tye) writes:
>In article <37 at consult.UUCP>, bob at consult.UUCP (Bob Willey) writes:
>>
(Miscellaneous descriptions and anti-flame on Computone deleted - wjv)

>I also found that the pass-thru print function can't handle a large amount
>of data especially with lots of terminals doing the same thing. One customer
>was losing parts of their print. Nobody at Computone could explain why it
>was happening. Replacing the board with a Corollary and in another instance
>a Digiboard solved that problem as well.

I have a site with 5 Computones, and all new boards being installed are
Anvils. Not that we had major problems with the Computones, but found the
Anvils were doing the job better and faster at the same transmission speeds.
I had some problems with Computone printing but finally got the
system working okay, but it may not solve your problem.

I have one AT style box with 7 users attached.  Four are PC's running terminal
emulation software with attached printers.  I was losing data on longer
printouts, but finally got it fixed (at least I haven't had any complaints
lately).

I had to set the pri parameters in the /atx subdirectory to make sure that it
looked at the keyboard often enough.  Set the time spent printing in
relationship to looking at keystrokes too high and you seem to lose the
handshake.  The terminal sends x-off but the host misses it.

I also had to set the buffer threshold in the terminal emulation.  It had
choices of none (as a terminal would) 64 bytes, 128, or 1024.  I seem to
remember that the 64 bytes xon/xoff threshold seemed to work fine.  If you are
running pure terminals that might be the problem - no buffering and swallowing
the printer handshake.

-- 
Bill Vermillion - UUCP: {uiucuxc,hoptoad,petsd}!peora!rtmvax!bilver!bill
                      : bill at bilver.UUCP