DataMuseum.dk

top - metrics - download

    Length: 3409 (0xd51)
    Types: TextFile
    Names: »ck9ker.bwr«

└─⟦9ae75bfbd⟧ Bits:30007242 EUUGD3: Starter Kit
    └─⟦3da672b63⟧ »EurOpenD3/misc/kermit.tar.Z« 
        └─⟦126dd3ca0⟧ 
            └─⟦this⟧ »ck9ker.bwr« 

TextFile

Date: Mon 22 Feb 88 00:28:49-PST
From: Bob Larson <BLARSON@ECLA.USC.EDU>
Subject: Os9/68k C-Kermit 4E Bugs & Problems

Known problems and limitations of ckermit 4E(070) for os9/68k

The alarm faking routines for the script and dial facilities is
incomplete.  I have not had problems with them in practice.

Timouts of 8 hours or greater are not supported.

I've noticed that a timeout occurs after recieving about 500 bytes via
a get command.

-------

One other note about the microcom modem: the single-character enter
command mode should be disabled (or at least changed from ^A).  Use
the SE0 command to do this.

More os9/68k C-KERMIT Bugs and Fixes.....

[Why do bugs always show themselfs at inconvenient moments?]

The midnight timeout problem was much worse than I expected, and I found
a better solution to the problem.  (Timeouts are now limited to 8 hours or
less.)

The parity stripping code had been moved to the tio code, and I didn't notice.
My new version now does parity stripping as required.

I made a couple of mistakes in the microcom dialer which caused it to fail
in extra long responce mode.

The first two arguments of memcpy should be cast to (char *).  I realize a
#define will be needed on systems that don't have memcpy.  (4.3bsd has bcopy
which just has the first two arguments reversed.)

I fixed a bug in ckudia.c of a possible array out of bounds assignement.

Bug and fix for osk C-kermit:

Symptom: set file commands cause address errors

Fix: remove the "int z;" declaration from the switch statement
following case XYFILE: in ckuus3.c.  This is apparently a compiler
bug.  (z is declared at the top of the routine and a separate
variable is not needed.)

Some of the variables in cmcmai are declared without type (other than
the unspecified default int.)  Shouldn't the ; following the cmask and
fmask declarations be changed to , to match the style of the rest of the
file?

-------

22-Mar-88 11:02:34-EST,1460;000000000001
Mail-From: SY.CHRISTINE created at 22-Mar-88 11:02:29
Date: Tue 22 Mar 88 11:02:29-EST
From: Christine M Gianone <SY.CHRISTINE@CU20B.COLUMBIA.EDU>
Subject: [Bob Larson <BLARSON@ecla.usc.edu>: os9/68k ckermit timeout problem understood]
To: sy.fdc@CU20B.COLUMBIA.EDU
Message-ID: <12384382062.192.SY.CHRISTINE@CU20B.COLUMBIA.EDU>

FYI.....
                ---------------

Return-Path: <BLARSON@ecla.usc.edu>
Received: from columbia.edu by CU20B.COLUMBIA.EDU with TCP; Tue 22 Mar 88 03:59:36-EST
Received: from ECLA.USC.EDU by columbia.edu (5.54/1.14) 
	id AA00202; Tue, 22 Mar 88 03:56:59 EST
Date: Mon 21 Mar 88 22:05:26-PST
From: Bob Larson <BLARSON@ecla.usc.edu>
Subject: os9/68k ckermit timeout problem understood
To: info-kermit@cu20b.columbia.edu
Message-Id: <12384273372.37.BLARSON@ECLA.USC.EDU>

[It seems that cu20b.columbia.edu is no longer in our host table.
Did it get dropped from the sri-nic table or is it just to many hosts
for our tops-20 systems fixed table again?]

After watching numerous file transfers, I think a finaly understand the
timeouts I mentioned in the ck9ker.bwr file.  My hard disk is very fragmented,
and apperently it takes long enough to allocate another extent to a file
to exceed the default 7 second timeout.  As the fragmentation gets worse,
the extents get smaller making the problem occur more often.  Fix: clean
up the disk, get a faster disk, or increase the timeout.

------------------------------