The Risks Digest

Forum on Risks to the Public in Computers and Related Systems

ACM Committee on Computers and Public Policy, Peter G. Neumann, moderator

Volume 5 Issue 65

Wednesday 25 November 1987

Contents

o Mariner I and computer folklore
Jon Jacky
Jim Horning
o Computer-controlled train runs red light
Jon Jacky
o Addressable CATV information
Ted Kekatos
o A new legal first in Britain...
Gligor Tashkovich
o The rm * controversy in unix.wizards
Charles Shub
o Info on RISKS (comp.risks)

Mariner I and computer folklore

Jon Jacky <jon@june.cs.washington.edu>
Tue, 24 Nov 87 22:13:08 PST
Mark Brader asks what really happened to Mariner 1, the Venus probe that had
to be blown up when it flew off course shortly after launch.  Some versions
of the story blame a missing hyphen, others blame a period substituted for a
comma.

I looked into this about a year ago.  I lost the trails of both versions
without finding a common ancestor.  Here is what I found out.  I hope some
reader can help.  The anecdote is told so often that someone really ought to
settle this once and for all.

  New York Times, July 23, 1962, p. 1 col. 2:  Atlas carrying Mariner I goes
  off course, destroyed by range safety officers

  New York Times, July 28, 1962, p. 1 col. 4: NASA, USAF, JPL announce Mariner
  I lost because flight control computer generated incorrect steering commands.
  Problem described as a "missing hyphen."

  New York Times, Aug 2, 1962, p. 24 col 5: Letter to the editor about Mariner
  I, calls for better computer programming practices.

  Mariner I loss attributed to substitution of period for comma in FORTRAN
  program: Henry S. Tropp, "FORTRAN Anecdotes," ANNALS OF THE HISTORY OF
  COMPUTING, Vol 6, No. 1, Jan 1984 pps. 61,62.    Tropp merely cites Jim
  Horning in ACM SOFTWARE ENGINEERING NOTES, 4(4) Oct. 1979 p. 6, who cites in 
  turn G.J. Myers, SOFTWARE RELIABILITY: PRINCIPLES AND PRACTICES, New York,
  John Wiley, 1976, p. 275.  

It looks like the missing hyphen version is much older.  I haven't been able
to trace the period-for-comma version to a printed source before Myers.
Still, I am not ready to accept the hyphen version as authoritative. I don't
have a copy of the NY TIMES story - I had to make notes from a microfilm
reader - but I recall that it seemed a bit confused, as if the reporter did
not quite follow the explanation he was given.  Also, I seem to recall
hearing the period-for-comma version long before Myers, when I was in
college around 1970.  Can anyone else offer an older citation?

A few leads I never followed up: obviously, Myers himself could to be
contacted to learn where he got the story.  I called IBM, credited in Myers'
book as his place of employment.  IBM said Myers had left some years ago and
they had no forwarding address.

Also, RISKS Volume 1 number 2, 28-Aug-1985, included a posting from Nicholas
Spies (Nicholas.Spies@CMU-CS-H.ARPA), in which he mentioned a memo about the
incident which his father had seen at the time.  No details were given in
that posting.  Nicholas, are you still there?  Can you help?

I think this matter would make an interesting case study for a folklorist. It
certainly has a lot of the aspects of the kind of urban folklore retold in the
book THE CHOKING DOBERMAN or an FOAF story ("this happened to a friend of a
friend").  In this case however, the tales are based on a real event in the
fairly recent past, so it should still be possible to find out what actually
happened.

It is interesting to note how a single incident gave rise to at least two
incompatible versions.  They now have an independent life - the RISKS index
in ACM SOFTWARE ENGINEERING NOTES, 12(1) Jan 1987 p. 23 cites both versions
as if they were two separate events.  The versions continue to fracture into
increasingly garbled variants.  The announcement for COMPASS '88 in
RISKS-5.62 said "a rocket to Mars had to be destroyed...".  The index in SEN
also mentions "Mariner 18 - aborted due to missing NOT in program".  It is
not clear where this comes from; possibly another Mariner 1 mutation, or
maybe it is supposed to be Mariner 8, which my ILLUSTRATED ENCYCLOPEDIA OF
SPACE TECHNOLOGY (by Kenneth Gatland, Harmony, 1984) says "was lost during
launch."  Whatever, there was no Mariner 18 - the last in the Mariner series
was 10, a 1973 Venus flyby.
                                          - Jonathan Jacky
          [Jon, MANY THANKS. PGN]


Mariner/Annals [A little duplication and a little more clarification]

Jim Horning <horning@src.DEC.COM>
Wed, 25 Nov 87 13:14:04 pst
The reference to ANNALS OF THE HISTORY OF COMPUTING, vol. 6, no. 1, should
be to page 61, not 6. However, it sheds little additional light: It quotes
my note in SEN October 1979, and my reference [3], G. J. Meyers, SOFTWARE
RELIABILITY: PRINCIPLES AND PRACTICES, John Wiley, 1976, p. 275. Meyers
doesn't cite his source, and I have never been able to get independent
confirmation.
                                      Jim H.


Computer-controlled train runs red light

Jon Jacky <jon@june.cs.washington.edu>
Tue, 24 Nov 87 22:16:54 PST
From IEEE INSTITUTE, Dec. 1987, p. 8:

CHIPS TOO UNRELIABLE FOR TRAINS, SAY ENGINEERS by Gadi Kaplan

"..This was one of the main conclusions at the Symposium on Microprocessors
in Rail Transit, held in Pittsburgh on Sept. 14-16 by the Rail Systems Center
of Carnegie-Mellon University's Mellon Institute. ...

Technical experts agree that microprocessor-based systems are more flexible
in operation and much better at monitoring and fault diagnosis than the
relay-based systems they typically replace. ...

Symposium participants expressed concern, however, about the probablity
of failure of the microprocessor in an unsafe way as a result of inadequate
verification of its software.  A case in point was the failure, in February
1986, of a four-car train operated by the Washington Metropolitan Area Transit
Authority (WMATA) to stop at a red signal. ... "The failure could not be
replicated with the same cars at the same location under any condition with
.. prolonged field and laboratory testing," (a WMATA official) reported...

However, a more postive view was expressed by panelists from ... suppliers
of microprocessor-based systems for rail transport.  These panelists said
they were confident their software, which required years to develop, at
extensive costs, was verifiable and reliable."

(End of excerpts from IEEE INSTITUTE)
                                                  - Jonathan Jacky


Addressable CATV systems

<ihnp4!ihuxv!tedk@ucbvax.Berkeley.EDU>
Wed, 25 Nov 87 07:39:33 PST
In my town, Oak Park, we have CATV provided by Cablevision of Oak Park. The
CATV control boxes have a serial number which is recorded (and phoned in to
the computer center) by the installer.

The digital signal broadcasted from the computer center (within the cable
company) provides the boxes with the date and time.  Niffy feature,
localized time base for all devices.  I have a button on my box for "display
time" which is displayed at the top of my screen.

But most importantly the digital signal transmits an individually addressed
(packet?) for each customer that provides a "matrix" of what each channel on
the box vectors to from the cable. I have noticed that the order of the
channels on the cable (itself) are different than what you see when you get
with the CATV box. The "Un-Authorized" channels, such as Playboy and HBO,
are _replaced_ with local cable guide (rather than the scambled signal and
sound). The CATV box stores the matrix even if un-plugged from power. When
the installer plugged in the box for the first time, all the channels where
un-authorized.

When I call the cable company for a "pay-per-view", they update the matrix
in my box to allow me to watch the program. The Matrix software in the box
might even have "HOW LONG" information in it.

Now, How do I get the localized time base to keep my Microware oven clock on
time ????? :-)

Ted G. Kekatos, AT&T Bell Laboratories, Indian Hill South, IX-1F-460
Naperville & Wheaton Roads - Naperville, Illinois. 60566 USA
backbone!ihnp4!ihuxv!tedk


A new legal first in Britain...

Gligor Tashkovich <gligor%lerouf.DEC@decwrl.dec.com>
25 Nov 87 20:15
I heard somewhere that Britain is experiencing a new legal first:

Apparently, a computer consultant is on trial there and is charged with
criminal damage by planting "logic bombs" in his clients' software.

Does anyone else have more information?


the rm * controversy in unix.wizards

Charles Shub <cdash@boulder.Colorado.EDU>
Wed, 25 Nov 87 09:54:27 MST
Yesterday, I got bit by rm [REMOVE]. I was remotely logged in to a system
over a network and had created a bunch of temp files. to delete them, I
naturally typed in "rm t*" only the %$*#&^#@ network managed to drop the "t"
and you all know what happened then. It wasn't too bad because with the
archiving we do it was only 2 hours to get them back. Of course yesterday's
changes got lost and had to be redone. The point is that there are two
things a command interface could do:
  1) protect us from our own stupidity (I'm not convinced it should),
  2) protect us from "extended system" errors like dropping a character,
but I'm not sure how you separate the two.

cdash   aka cdash@boulder.colorado.edu    aka ...hao!boulder!cdash
    aka ...nbires!boulder!cdash       aka  (303) 593-3492

Please report problems with the web pages to the maintainer