[comp.unix.xenix] SCO gyrations

shevett@mccc.UUCP (Dave Shevett) (09/15/89)

This article has been tumbling around in my head for several days now, but
has just reached the point where I GOTTA get some answers:

Configuration: SCO/Xenix v2.3.1 no updates
			Wang PC-381-7 (16mghz, 4meg RAM, 2 CDC 42 megs, Emulex 8port
			serial mux, 2 normal ser, 1 parallel, Oki 2410 on tty2a)
			Truckloads of applications (Foxbase, SCO Pro, WordPerfect,
			etc)

Problem 1 - A while back someone mentioned the slowness of the Xenix
printer spooler.  I assume other people have had this problem.  One of my
printers will do a 1/2 second pause between lines when printing.  The
banner page comes out fine, but the report (say from Realworld accounting)
takes forever. Suggestions? NOTE - This only happens on the parallel port.

Related prob (prob 1a) - I'm ready to chuck the whole spooler package and
go to direct /dev/tty2a prints.  In tinkering with the lp scheduler, trying to
get various configurations to work, the scheduler invariably decides NOT
to send anything to the printer.  I enable printer1, and queue something
(ala ll|lpr).  Hmmm, ho hum.  Gee, nothing's printing.  lpstat -t shows the
scheduler running, the printer enabled, AND all my jobs.
The job is queued, but never prints.  I try lpshut and lpsched to restart
the job, I try killing off the queue and resubmitting - nothing.  However,
an lx>/dev/tty2a DOES print (!) - Do I have to goose the lp scheduler to get
started?  If I reboot the machine, as soon as 'init' restarts the
scheduler, BING - all waiting print jobs spew out.  This is VERY
frustrating.

Problem 2 - Recently my machine has refused to acknowledge that COM2:
exists.  The opening configuration screen only shows serial port 1 with
nports=1.  I can do all my fun stuff on that port, but trying to enable
/dev/tty2a comes up with the perpetually dying getty.  Occasionally the
system will come up with com2: enabled (both show in the configuration),
but those situations are few and far between.

Problem 3 - I'm running an Emulex DCP/MUXi 8 port intelligent card as
tty61-tty68, and the system runs fairly well.  However, in Foxbase, there
is a 1 second delay when a field is selected.  For most functions, there is
no delay - ONLY when an @ x,x get mfname is executed is there this 1 second
delay.  VERY tedious when filling out a screen full of them.  Emulex has
been helpful, but baffled.  They had their driver written by INETCO, and
they (inetco) say all settings are correct.  (they had me fax them the stty
-a settings out of Foxbase and while running it).  I told Emulex several
people on the net mentioned other boards had this problem when they were
first released, and he'll pass it along.  Can anyone from Computone or
Arnet TELL ME WHAT CAUSES THIS?  Then I can tell Inetco how to get the
*&^%$ thing working.

Problem 4 - This is more a request for information - I've found a source
for DTK Ethernet cards for around $140.  This *sounds* great, but lemme
tell you what I want to do.  I will shortly have an SCO system (same as
above) home for developmental work.  BUT, I can't shell out the bucks for
a multiport card.  I will also have SCO TCP/IP, SCO NFS, and SCO XENIX-NET
for evaluation.  I'd like to hook up my 3 pc's at home (AT class machines)
as terminals via TCP/IP into the Xenix box - CAN THIS WORK?  Will something
like ka9q (which I have) allow an rlogin to the Xenix box via the ethernet
link?  Question 2 - If it does, how does the terminal emulation work?  Will
I just be an ANSI terminal?

Any information back on this will be helpful.  Our SCO contract is in the
works, and I have NO ONE to talk to about these things.  I'm desperately
trying to get xenix accepted within our company as 'the only way to go',
(I'm competing against a System/36 - can you believe it?), but it's
difficult when no one has the answers.

+----------------------------------------------+------------------------\
| Dave Shevett   | Dupli-Fax Technical Support | "The shortest distance |
| W. Trenton, NJ |      Burlington, NJ 		   | between two puns is    |
| (609) 883-6734 |      (609) 387-8700		   | a straight line..."    |
+----------------------------------------------+       - Doc Webster    |
\ Usenet: princeton!mccc!shevett     Inet: shevett@mccc.UUCP >----------/

bob@consult.UUCP (Bob Willey) (09/16/89)

Your first problem can be resolved with patch xnx115 available
from SCO.  It is for Slow Parallel Printing Fix for
SCO Xenix 386AT 2.3.0/2.3.1 .
Regarding your other printer problems, sounds like lpsched is not
running.  Check via ps -ef to see that it is active when you are
initiating printing.
Good Luck.

-- 
.. Computer Consulting Service     ..      Bob Willey        ..
.. P.O. Drawer 1690                ..    uunet!consult!bob   ..
.. Easton, Maryland  21601         ..    (301) 820-4670      ..
...............................................................

jdeitch@jadpc.cts.com (Jim Deitch) (09/17/89)

In article <482@mccc.UUCP>, shevett@mccc.UUCP (Dave Shevett) writes:
..... deleted
> 
> Problem 2 - Recently my machine has refused to acknowledge that COM2:
> exists.  The opening configuration screen only shows serial port 1 with
> nports=1.  I can do all my fun stuff on that port, but trying to enable
> /dev/tty2a comes up with the perpetually dying getty.  Occasionally the
> system will come up with com2: enabled (both show in the configuration),
> but those situations are few and far between.
> 
I have had this happen to me a couple of times.  I found it to be that the
COM2 port was not being reset properly when doing a shutdown.  Try shutting
the machine off and turning it back on.  If this seems to bring the port
up every time, it could be the uart or support chips are bad.  Try a different
card and see if it still reacts properly.  Also, check for conflicts, interupt
and i/o adresses that could be causing a conflict.

Hope this helps some.

Jim

Now, if only I could find a way to debug my kernel.....