[comp.dcom.lans] tn3270 w/ suns problem

rwc@mhinfo.UUCP (Ralph Callaway) (12/25/89)

Thanks for the replies regarding problems with the tn3270 code.

We are now using the V 4.1.1 code for the tn3270 emulator on sun
worksations (sun3 and sun4).  Everything seems to be working 100%
except for one thing.  I was puzzled since in the ANNOUNCE message
describing fixes, this was mentioned as fixed...

The problem is that when the user goes into an unformatted (empty)
CICS screen the screen properly blanks but cpu consumption goes to
100% and the cursor moves rapidly through the right hand margin.  The
only way that we have been able to recover is to kill the process.

Has anyone else experienced this kind of behavior or have suggestions
on where in the code we should start to look?

If possible I would appreciate responses by mail.  I will format
the answers/solutions/suggestions and post in the news group.

	!uunet!mhinfo!rwc


Ralph Callaway

P.S.  One last question, does anyone have a fix to display the status
line.  Specifically, the users want to see the line showing things
like the insert toggle mode.

pwl@tc.fluke.COM (Paul Lutt) (01/03/90)

In article <20@mhinfo.UUCP> rwc@mhinfo.UUCP (Ralph Callaway) writes:
>The problem is that when the user goes into an unformatted (empty)
>CICS screen the screen properly blanks but cpu consumption goes to
>100% and the cursor moves rapidly through the right hand margin.  The
>only way that we have been able to recover is to kill the process.
>
>Has anyone else experienced this kind of behavior or have suggestions
>on where in the code we should start to look?
> Ralph Callaway

I know Ralph requested email on the subject, but I feel compelled to
post a note here.  I have also seen this problem, starting with version
4.1 of the tn3270 program.  I hacked together a really gross fix for
our site and contacted Greg Minshall about this problem.  He replied
that version 4.1.1 had fixed some problems with CICS screens.  I
obtained a copy of version 4.1.1, but it still didn't work correctly.
I sent Greg copies of some network traces that show the traffic between
the tn3270 host and the IBM system, in hopes that he might be able to
duplicate the problem.

I would like to see these problems fixed by Greg or someone in his
group.  My patchs seem to work, but I don't trust them for general
distribution, since my understanding of how the tn3270 package works is
incomplete.

I would encourage anyone who is having similar problems either post
about them here or send a note to Ralph (rwc@mhinfo.UUCP), who has
volunteered to collect them and then post a summary to the net.

Let me end by saying I do appreciate all the work that Greg Minshall
has done to provide this valuable package.  I just hope that we can get
the last few quirks taken care of.
--
	Paul Lutt
Domain:	pwl@tc.fluke.COM
 Voice:	+1 206 356 5059
  UUCP:	sun!fluke!pwl
 Snail:	John Fluke Mfg. Co. / P.O. Box 9090 / Everett, WA 98206-9090