Human-Nets-Request%rutgers@brl-bmd.UUCP (Human-Nets-Request@rutgers) (01/10/84)
HUMAN-NETS Digest Tuesday, 10 Jan 1984 Volume 7 : Issue 6
Today's Topics:
Adminstrivia - Testimony of Willis Ware for Xerox folk,
Input Devices - Keyboards &
The Original Question &
DVORAKs,
Computers and the Law - Big Computer is Watching you (2 msgs),
Computer Networks - Networks, Networks, Everywhere (2 msgs),
Computers on TV - Whiz kids
----------------------------------------------------------------------
Date: 5 Jan 84 15:16:25 PST
From: Charles <mcgrew@rutgers>
Subject: Testimony of Willis Ware for Xerox folk
Willis Ware's testimony (announced in a previous digest) is
available for people with access to Parc in
[Ivy]<Registrar>Testimony.Txt
Charles
------------------------------
Date: 7 January 1984 00:12 EST
From: Jacob Moskowitz <JMSK @ MIT-MC>
Subject: Keyboards
Date: Thu, 5 Jan 84 19:27:24 pst
From: decwrl!qubix!msc at Berkeley (Mark Callow)
To: HUMAN-NETS
The phone keyboard was designed before small electronic
calculators existed.
What about mechanical adding machines ? didn't they have the same
keypad layout as modern calculators ?
------------------------------
Date: 7 January 1984 01:29 est
From: Makey.DODCSC at MIT-MULTICS
Subject: The Keyboard as an Output Device
Thanks to everyone who responded with info on Dvorak keyboards,
and especially to Chris Jones for the layout diagram. Moving back
toward the subject that started all of this (input devices in general
and when/why/how you would use them), Greg Davidson (in Human-Nets,
vol. 6 no. 80) said:
I believe that the question of how to support non-standard
keyboards, such as DSK keyboards and chord keyboards, has a
simple answer: Make a standard interface which is
independent of which one is used. People should be able to
plug their favorite keyboard into any system.
I have a similar answer for the support of various pointing
devices, including mice, tablets with pens, tablets with
pucks, touch screeens and light pens. A standard port on
terminals and workstations should accommodate any such
system, even if something else is built in.
He is basically correct about the need for a standard interface.
But, which standard? Will all input devices have the same interface
or will there be a different one for each type (i.e., keyboard, mouse,
joystick, etc.) of device? The computer industry seems to recognize
the value of standardization so I think it is a question of how and
not whether the standardization will occur.
Just a couple of questions: Is a standardized plug-in interface
enough? For example, how do you interface a one-button mouse to a
system that really wants a three-button mouse? When you plug your own
keyboard with its own arrangement of special function keys into a
system, how do you know which keys do which functions?
I have an interesting solution to this last problem: Have the
design on each key change in real time under host (or local) control
so it always displays its current function. For example, when editing
the function keys would show what editing functions they do, but when
you run the mail program the keys automatically display mailing
functions. If you want to go all out, the alphabetic keys would
normally display lowercase letters, but when you press the "shift"
key, they would all immediately display uppercase letters.
The difficult part of doing this is the per-key displays. LED
arrays or LCDs would work, but I suspect that these would add
significantly to the mass of each key, thus affecting the "feel" of
the keyboard (I don't think I would like it). Some form of visual
projection system, where the top of each key is a tiny rear-projection
screen, could eliminate the key mass problem, but the rest of the
apparatus is likely to be bulky and I'm not sure how well it would
work in high ambient light situations.
Any other ideas?
:: Jeff Makey
------------------------------
Date: 7 January 1984 03:11 EST
From: Jerry E. Pournelle <POURNE @ MIT-MC>
Subject: DVORAK Keyboards
In response to the inquiry on where I got my figures: they're
hazily recalled from the days when I worked for August Dvorak
back at the University of Washington. He was no longer doing
keyboard studies, but he had never given up, and every now and
then would haul out his studies, movies, and recordings (actual
old records; no tape recorders in the 30's, or if there were, he
didn't have one) of the sounds of people typing (faster under
his keyboard).
The problems of going back to the standard when you
learn Dvorak are reported from personal experience: we had a
Dvorak typer at UW and I worked with it for a while just to see;
I didn't notice a lot of improvement in typings speed, and had a
HELL of a time returning to Qwerty when I decided to give up.
------------------------------
Date: 7 January 1984 02:58 EST
From: Jerry E. Pournelle <POURNE @ MIT-MC>
Subject: the IRS welcomes you to 1984 ... (a true story)
Wonderful news. In my case there are at least twenty different
names under which I get magazines, junk mail, etc., including
some really unlikely spellings like Dournelle and in one case
Dear Mr. pournelle Photography, you may have won a prize... thnk
of what your neighbors in North Hollywood will think when they
see y ou, Mr. Photography, in a new LeBaron..." (My son had
some business cards printed once listing this as "Pournelle
Photography." I get mail to Dear Dr. PhD, and I once filled out
a consumer survey form for Dr. PhD, listing his income as, as I
recall, "over $50,000"; that ought to get the IRS excited...
Ye gods.
------------------------------
Date: 7 January 1984 19:04 EST
From: Robert Elton Maas <REM @ MIT-MC>
Subject: Thoughtcrime
If a database includes only information that is based on solid
evidence, and continually/recurrently rechecks information cleans up
typograhpical errors that creep in and deletes any data that turns out
not to be correct or which is based solely on opinion rather than
fact, then I don't think we have much to worry about. On the other
hand, if a database isn't cleaned up, if it has garbage that came from
unknown source at unknown time which has never been substantiated or
checked, then it should be easy for the citizenry to flood the
database with so much random cruft that the database becomes obviously
worthless to everyone involved even if there's enough mass-storage to
keep it all.
Well, now I'm probably on the FBI's list as somebody who might
sabotage their database by feeding it more GIGO than it can stomach.
[puns deliberate]
------------------------------
Date: 9 Jan 1984 1311-PST
From: Chuck McManis <MCMANIS@USC-ECLC>
Subject: Networks
Of course following such a glowing report on the joys of DECnet by
John Covert I find it necessary to mention that I like decnet between
my RSX and VMS system but I hate trying to use it between my TOPS-20
and either RSX or VMS. Seems the old 2060XE got left off the
distibution list of compatiblity. Are you listening DEC ?
--Chuck
------------------------------
Date: Mon 9 Jan 84 15:52:34-MST
From: Walt <Haas@UTAH-20.ARPA>
Subject: Yet More Networks
In addition to the networks previously described, there are five
public data networks actively serving the US and more in the works.
The five national PDNs are all common carriers, like Greyhound - that
is, anybody whos pays the fare can use them. They all provide an X.25
interface, which gives a virtual circuit service - there is as yet no
international standard for mail or FTP. All provide a virtual
terminal capability via the X.3/X.29 PAD standards. They all compete
vigorously for business, and I'm sure I'll hear about it immediately
if I have left out anybody's capability. Here (in alphabetical order)
are the five established PDNs:
o ADP Autonet
175 Jackson Plaza
Ann Arbor, MI 48106
(313) 769-6800
Besides the US, has satellite links to London, England and Delft,
The Netherlands. Maximum internal speed is 9600 bps. Nodes are
PDP-11s with KMC-11 front end microprocessors. Internal protocol
was described to me as derivative of the old ARPAnet protocols.
o CompuServe Incorporated
Network Services Division
5000 Arlington Centre Blvd.
P.O. Box 20212
Columbus, OH 43220
(614) 457-8600
Internal speeds to 56k bps. Nodes are PDP-11s with 6809
microprocessor front ends. Internal protocol is DDCMP.
o GTE Telenet Communications Corp.
8229 Boone Boulevard
Vienna, VA 22180
(703) 442-1000
Internal speeds to 56k bps. Nodes are arrays of 6502s in a
redundant, load sharing configuration. Internal protocol
conforms to CCITT Recommendation X.75. Supports automatic
recovery of virtual circuit when a node fails during a call.
Built by some of the folks from BBN who built the ARPAnet
originally. Provides a mail service called Telemail.
o Tymnet, Inc.
2710 Orchard Parkway
San Jose, CA 95134
(408) 946-4900
Internal speeds to 56k bps. Nodes are arrays of "Tymnet Engines"
in a redundant, load sharing configuration. The Tymnet Engine
is a Tymnet-built 32-bit processor derived from the Interdata 732,
re-engineered for extremely high MTBF. Internal protocol is a
unique Tymnet design which repacketizes inside the network and
does flow control at the byte level, like TCP. Supports automatic
recovery of virtual circuit when a node fails during a call.
Provides a mail service called OnTyme.
o Uninet
United Telecom Communications, Inc.
2525 Washington
Kansas City, MO 64108
(816) 221-2444
Internal speeds to 56k bps. Nodes are Modcomp 7830s. Internal
protocol is a Uninet-designed virtual circuit protocol, on top
of HDLC.
In addition there is, of course, the new AT&T offering, NET/1000.
Nodes consist of arrays of VAXen with a Series/I for line handling.
They see the function of their network as storing information, rather
than just forwarding it like the other networks. The internal
protocol is X.25, but they don't support an X.25 user interface! (No,
I don't know why). For further information, call Mr. John M. Finn,
their San Francisco account executive at (415) 452-7292.
Graphic Scanning and Computer Sciences Corp. are in the process of
spinning off their internal networks, as GraphNet and InfoNet
respectively I believe. There will probably be X.25 interfaces, if
they don't exist already.
GE Information Services Company has an internal network called
MARK*NET. There is not as yet an X.25 interface to it.
And, how could I forget, the State of Utah boasts its own Public Data
Network! It is called ComWest and is being spun off by Blue
Cross/Blue Shield of Utah, which needed a good way to get claims data
from places like Panguitch, Utah up to Salt Lake City. The internal
circuits are leased from Mountain Bell (no, they're not barbed wire,
skeptics) and run up to 9600 bps. Nodes are Dynatech Packet
Technology Multi-Switch.25 packet switches, which are based on the Z80
micro. There are several sites besides BC/BS, one of them being the
University of Utah DECSYSTEM-20.
Outside the US, there are public data networks operating in about
forty foreign countries, basically the ones that are industrialized.
We have a user who logs in regularly from Stockholm via the Swedish
PDN <-> Telenet <-> ComWest. He says he gets good response.
Cheers -- Walt
------------------------------
Date: 8 January 1984 00:17 EST
From: Robert Elton Maas <REM @ MIT-MC>
Subject: Whiz kids - remote intrusion into nuclear-war
Subject: scenerio database
Tonight's episode of Whiz Kids is awfully scary, in more ways than
one. Start with some "famous hacker" with nickname/moniker "wrench"
who posts a message on lots of public/school bulletin boards giving
the dialup number and password (a very long pseudo-random-looking one,
I think they've been reading Human-Nets lately), which turns out to be
an NSA computer. Next, when Ritchie succeeds at figuring out how to
get past additional security safeguards, like if you don't precede
each normal character with a hex C2 prefix (mark-parity quotes;
erroneously referred to as a non-ASCII character) it hangs up
instantly, about ten NSA people swarm around his house, swarm into it
right after they shut off the circuit breakers, confiscate *all*
computer equipment and *all* magnetic and non-magnetic media they can
find, and arrest everyone in the room with the computer, arraigning
them as adults because their knowledge of NSA access methods means
they have knowledge of adults... I won't spoil the plot by telling
the surprise of 28 minutes into the program nor the additional
surprise of 41 minutes into the program, but it's almost as scary as
WarGames and more realistic.
Sigh, I can't resist, the bad guys gain control of the whole
communications satelite and microwave network that the telephone
company and TV networks use, and shuts the whole thing down in unison.
You scared? I am. Remember that simple high-school people with
blueboxes made a practice of gaining control of individual central
offices. I think they once shut down a whole section of Los Angeles as
a demonstration several years ago and threatend to shut down all of
California or somesuch, I forget the details, perhaps Lauren can
refresh my memory.
End with the gullability of people, who can be led far astray if they
don't check what people tell to them to find out if it's really the
truth.
I hope this particular episode is shown on reruns soon so lots of
people who missed it this time can see it. <Personal opinion of REM
who is still a bit shaken from watching it. Too bad it's not on PBS
where'd it'd get rerun three more times in the next week so everyone
can see it.>
------------------------------
End of HUMAN-NETS Digest
************************