info-kermit@ucbvax.ARPA (06/22/85)
From: Frank da Cruz <SY.FDC@CU20B.ARPA>
Info-Kermit Digest Fri, 21 Jun 1985 Volume 2 : Number 36
Departments:
ANNOUNCEMENTS -
Another New C-Kermit for Unix (4C(53)) and Macintosh (0.8(32))
New FTP server up on CU20B: CWD fixed
Okstate Dialup Repaired
C-KERMIT -
C-Kermit Very Slow on TRS-Xenix
C-Kermit Runs on Heurikon Mini-Box
Mapping Keys in Mac Kermit
MS-DOS KERMIT -
Generic MS-DOS Kermit Runs on the DG/1
Kermit-MS under Topview
----------------------------------------------------------------------
Date: Fri 21 Jun 85 20:00:27-EDT
From: Frank da Cruz <SY.FDC@CU20B.ARPA>
Subject: Another New C-Kermit for Unix (4C(53)) and Macintosh (0.8(32))
To: Info-Kermit@CU20B.ARPA
C-Kermit 4C(053) fixes a major problem for both Unix and the Macintosh:
it now does parity right. There were two areas in which problems showed
up:
1. Macintosh Kermit could not transfer files using mark or space parity.
Furthermore, when using any kind of parity other than none, occasional
incoming characters would be lost during terminal emulation. This was
because Mac Kermit was letting the serial i/o chip handle parity, and the
chip is simply not flexible enough to be relied upon.
2. Unix Kermit could not transmit odd parity correctly. If you "set parity
odd", it was equivalent to "set parity mark".
The Unix Kermit software parity generation function dopar() was fixed to
generate all kinds of parity correctly. Then Macintosh Kermit was changed
to use dopar() rather than its chip to generate parity.
Several other minor changes were made to Unix Kermit, in particular in the
determination of whether it is in local or remote mode. The changes are
detailed in KER:CKUKER.UPD and KER:CKMKER.UPD; known bugs are listed in
KER:CKUKER.BWR and KER:CKMKER.BWR (U for Unix, M for Macintosh). The
complete collection of C-Kermit files for Unix, VMS, and the Macintosh are
in KER:CK*.*.
These files are all accessible via anonymous Internet FTP from host CU20B.
A message about CU20B's FTP server follows. Also, the file KER:AANETW.HLP
has been updated to provide more complete information on access to CU20B's
Kermit distribution area.
------------------------------
Date: Thu 20 Jun 85 23:39:51-EDT
From: Ken Rossman <sy.Ken@CU20B.ARPA>
Subject: New FTP server up on CU20B: CWD fixed
To: Info-Kermit@CU20B.ARPA
Address: 715 Watson Labs, 612 W. 115th St, NY NY 10025
Phone: (212) 280-4876
CU20B is now running a new version of the CMU TOPS-20 FTP server. The
major enhancement which will be important to most users is one which
(indirectly) makes multiple file transfers easier in some cases.
One problem (particularly with originating Unix systems) was that an mget
would return a list of files with fully qualified TOPS-20 pathname
prefixes, which in turn would be the way those files were created on disk
(i.e. the Unix filenames would contain the entire TOPS-20 path prefix).
While this is merely annoying on Berkeley 4.2 systems, this is a real
problem with 4.1 systems (and perhaps others), which have shorter file
names and can lose some of the name information.
Due to the new enhancements to the FTP server, a CWD command may now be
issued to connect the user to the desired Kermit directory (with or without
a password), so that the file list that FTP retrieves for a multiple file
transfer will not contain the entire directory path (just the filenames).
If no password is supplied at the time the CWD command is issued, no "owner
access" is granted to the target directory, and files are only accessible
according to their individual file protections (in the case of the Kermit
directories, this translates to read-only access). If a password is
supplied, full owner access is granted to the target directory.
Many thanks to Vince Fuller of CMU for the new FTP server. /Ken
------------------------------
To: info-kermit@CU20B.ARPA
Subject: Okstate Dialup Repaired
Date: 11 Jun 85 18:14:52 CDT (Tue)
From: Mark Vasoll <vasoll%okstate.csnet@csnet-relay.arpa>
Some of you have been reporting problems with the dialup access that we
provide to the Kermit Distribution. The problem has been traced to a
flakey modem and the offending unit has been replaced with a new unit.
Please continue to direct problem reports and suggestions to one of the
"uucp-support" addresses.
UUCP: {cbosgd, ea, ihnp4, isucs1, mcvax, pesnta, uokvax}!okstate!uucp-support
ARPA: uucp-support%okstate.csnet@csnet-relay.arpa
Mark Vasoll
Department of Computing and Information Sciences
Oklahoma State University
UUCP: {cbosgd, ea, ihnp4, isucs1, mcvax, pesnta, uokvax}!okstate!vasoll
ARPA: vasoll%okstate.csnet@csnet-relay.arpa
------------------------------
Date: Mon, 17 Jun 85 09:38 MDT
From: RMark@DENVER.ARPA
Subject: C-Kermit very slow on TRS-Xenix
To: Info-Kermit@CU20B.ARPA
I just installed 4C(052) on TRS-XENIX and timed a file transfer to a
very lightly loaded VAX/VMS. With a 4800 baud connection, the transfer
averaged 13 chars/sec.
[Ed. -- Wow, does everybody else with TRS-Xenix have the same performance?]
------------------------------
Date: 18 Jun 85 08:28:10 PDT (Tuesday)
From: Cherry.Pasa@Xerox.ARPA
Subject: C-Kermit Runs on Heurikon Mini-Box
To: Info-Kermit@CU20B.ARPA
C-Kermit has been sucessfully installed on two Heurikon Mini-Boxes.
(68010, Unisoft System V 5B 4/85; HK68 CPU - 10 MHz)
only two minor problem areas:
1. Tabs were converted to spaces at some point along the transfer route.
2. the " \ " characters had to be removed in the initial Make line.
Recommendation:
Since there are so many entries for a "3" type installation, change the
name of sys3 to sys5. I found that some non-unix types had trouble with
this as they have System 5 Unix and all through the documentation Unix V
is the more common term used.
Note: Time to complete the make process 27 minutes.
Thanks to all for a fine job on the software.
------------------------------
Date: Thu, 20 Jun 85 11:22:09 EDT
From: Greg Lauer <glauer@srn-vax>
Subject: Mapping Keys in Mac Kermit
To: info-kermit@cu20b.arpa
I used the kermit key map program to put control characters on the Option key
and meta characters on the Command key. The problem is that Option-e,i,u,n
still act like option keys: they "expect" another character. Thus to send a
control-n requires typing option-n option-n. Otherwise, this seems like a
great program.
Greg
[Ed. - The lowercase vowels "aeiou" are treated specially in the translation
routine /usr/mac/ws/local/init0. init0 is also the routine which was doing the
shift-clover-1..9 stuff. Mac Kermit currently has a command for disabling
the shift-clover-1..9 keys, but adding one for the option-vowels would be a
lot harder. For now, this goes into the .BWR file.]
------------------------------
Date: Wed 12 Jun 85 14:08:13-CDT
From: Aaron Temin <CS.Temin@UTEXAS-20.ARPA>
Subject: Generic MS-DOS Kermit Runs on the DG/1
To: info-kermit@CU20B.ARPA
We have a copy of MSGENE.EXE running on our Data General/One. File transfers
work fine at 1200 baud through an external modem. Internal modem doesn't seen
to respond, and screen emulation, even with the vt100 ansi.sys installed,
doesn't seem to work. But this is entirely satisfactory until someone can
modify Kermit appropriately for the beast.
Hope this is helpful. Thanks.
Aaron Temin
------------------------------
From: Jim Gillogly <jim@rand-unix>
Date: 12 Jun 85 14:38:00 PDT (Wed)
Subject: Re: Kermit-MS under Topview?
Yes, MS-Kermit v. 2.27 will run under Topview. I haven't bothered with
a .PIF file, but it works fine in the background. For 2.27 on the IBM PC
I used 96K for the min memory, 96K for max, and 7K for system. I left
the interrupts at 00-FF, although I'm sure they don't all need to be swapped
(if any) ... keep it simple. Use "n" for "program writes to screen", since
writing to the screen and running in the background are linked. The
disadvantage of telling it "n" is that it intercepts everything, giving you
a "galumphing" scroll. However, "n" is necessary for background, and if
you're transferring long files it's worth it.
I used 2.27 in background mode to pull Webster's 2nd over while doing a
half-day's worth of real work on my PC in the foreground. I normally don't
use Topview because it's not compatible with the Tall Tree RAMdisk ... and
since I have 2.5 meg of RAMdisk, that's a big loss.
It also works with MS-Kermit v.2.28, which needs only 85K for min/max size.
Jim Gillogly (jim@rand-unix)
[Ed. - Thanks for the information. Since 2.28 does dynamic memory allocation,
it will probably have to be defined differently to TopView.]
------------------------------
End of Info-Kermit Digest
*************************
-------