The Risks Digest

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 10 Issue 47

Thursday 4 October 1990

Contents

o California DMV and their new computer (Cecil Lee
2)
o Report of Nat Semi clock chip flaw
Martyn Thomas
o BA 747-400 Engine Failure
Martyn Thomas
o Novel on corporate computer espionage
Philip Brewer
o CERT Advisory - NeXT systems
Edward DeHart
o Fair Information Principles
Jeff Johnson
o Television rating (nee universal listening) device
Tim Wood
o From under a Rock??? (Subliminal message lawsuits)
Ed Hall
o Operation Sun Devil invades the InterNet?
Ed Luke via Michael Packer via John M. Chapin
o Info on RISKS (comp.risks)

California DMV and their new computer

Master Of Darkness <clee@td2cad.intel.com>
Sat, 29 Sep 90 21:57:26 PDT
    I just received my registration notice from the DMV (Department
of Motor Vehicles).  This piece of paper shows the amount of money I
need to pay, so that I can drive my car for the next year in California.

    Imagine my surprise when I saw the amount: $2214.  An extremely big
surprise since I only paid ~$600 last year, when I purchased the car new.  The
reason for this new method of making money?

    This is the excuse from the clerks at the DMV.  Seems that they have
just started using a new computer program for billing purposes, apparently on
their new Tandem computers.  Surprise Surprise it seems that everybody who got
a recent registration statement, has an invalid amount.  I had to go to the
local office so that they could print up a new form with the correct amount.  I
don't know if registration statements were the only incorrect items produced.
They didn't volunteer the information and I didn't ask.

    Supposedly some people received bills for less then the correct amount.
I wonder if they will still get their registration tags when (if) the computer
notices the underpayment.  Will those who might have paid too much get refunds?

    BTW, the correct amount I'm supposed to pay?  $351.

          Cecil Lee, Intel Corp.

CLee@SC9.INTEL.COM or CLee%SC9%SC.INTEL.COM@RELAY.CS.NET
UUCP : {pur-ee,qantel,amdcad,oliveb,decwrl,hplabs}!intelca!mipos3!sc9!clee


California DMV Troubles

Master Of Darkness <clee@td2cad.intel.com>
Sun, 30 Sep 90 16:42:30 PDT
The following was an article in the Sept 30, 1990 issue of "The Argus."
One of the local papers in the SF Bay area.

DMV computer goof overbills car owners
By Mark van de Kamp (staff writer)

    Sham Dixit of Livermore was one of many California drivers who felt
they were being asked to pay too much when they got notices this week to renew
their motor vehicle registrations between now and November.  And they were
right.  The state Department of Motor Vehicles admitted Friday that it had made
a blunder which caused some drivers to be overbilled by hundreds, even
thousands of dollars.  In Dixit's case, he was asked to pay $2,832 for his 1987
Nissan Sentra.  It cost him $166 to register the car last year.  Likewise,
three members of a Pleasanton family were overbilled by $1,000 each.

    The DMV does not know how many vehicle registrations are involved.
There are 25 million registered vehicles in the state.  The agency said the
incorrect billing notices involve drivers whose vehicle registration was set to
expire Nov. 16, 18 and 20.  Most of the incorrect bills arrived in mailboxes
Thursday and Friday.  [...]

    The problem surfaced late Thursday when DMV offices started receiving
calls from motor vehicle owners asking why registration fees had changed
significantly from the previously year, the agency said.  "At first thought I'd
made a mistake.  Then I heard that the DMV screwed up.  Boy, did they ever,"
Dixit said.  "But I work with computers, so I know it must be a programming
error.  Computers are only as smart as the people who use them." [...]

          Cecil Lee, Intel Corp.                   [PGN Excerpting Service]

CLee@SC9.INTEL.COM or CLee%SC9%SC.INTEL.COM@RELAY.CS.NET
UUCP : {pur-ee,qantel,amdcad,oliveb,decwrl,hplabs}!intelca!mipos3!sc9!clee


Report of Nat Semi clock chip flaw

Martyn Thomas <mct@praxis.co.uk>
Thu, 4 Oct 90 12:38:35 BST
Electronics Times (4 Oct, front page) reports that National Semiconductor's
real-time clock chip (part number MM58274B) "has a tendency to switch from a
24hr clock to a 12hr clock when subjected to electronic noise ..".

Two examples are given of problems allegedly caused by the chip.

"... the chip caused the time clock in a financial system to skip from
Thursday to Saturday, leaving employees without paychecks".

"It has also caused problems for the United Nations Atomic Energy Agency
which uses the chip in a televised security system for guarding nuclear fuel
.... ".

Martyn Thomas, Praxis (Software Engineers), 20 Manvers Street, Bath BA1 1PX UK.
Tel:    +44-225-444700.   Email:   mct@praxis.co.uk


BA 747-400 Engine Failure

Martyn Thomas <mct@praxis.co.uk>
Wed, 3 Oct 90 15:21:58 BST
Flight International (3-9 October) reports that a British Airways Boeing
747-400's No 1 engine electronic controls failed on takeoff at London
Heathrow causing the engine to shut down.

The crew [two pilots, there is no flight engineer] reported the status
message "engine controls" and asked their technical support staff, by radio,
for advice. They were told "You've obviously lost control of that engine.
It's a FADEC failure" [FADEC = Full Authority Digital Engine Controller].

BA says that the problem was a spurious signal from the electronic "thrust
reverse resolver". If so, the early diagnosis of FADEC failure could be
wrong. There has been a number of instances of spurious signals causing
747-400 engines to throttle back or shut down, according to Flight [ This
may be a reference to the earlier reports of spurious signals from flap and
gear sensors, reported in an earlier RISKS].

Flight adds that FADEC failure is extremely unusual.

Martyn Thomas, Chairman, Praxis plc. Software Engineers.
Tel:    +44-225-444700.   Email:   mct@praxis.co.uk


Equinox on the A320: Programme summary

1 Oct 1990 14:29:40-BST
Below is a summary of the Channel 4 (UK TV) programme on the A320
transmitted at 7pm Sunday 30 September.  I took notes during the
programme but I may have got some details wrong.

Equinox asked an independent air accident investigator named Ray Davis to
examine the report on the Habsheim crash where an A320 being flown in a display
made a slow pass over the runway and could not pull up in time to avoid the
trees at the end of the airport.  He made four major findings which were put to
the Chief engineer (I think) at Airbus Industries.

1: A 4 second discrepancy between the Cockpit Voice Recorder, the
   Aircraft Data Recorder and the Tower Voice Recorder.  Sorry but I
   cannot remember the Airbus reply to this.

2: There was no record of the impact with the trees in the flight
   data.  This might be expected in (say) a collision with a mountain,
   but the recorders should have been able to operate until the
   aircraft disintegrated.  Any crash which could be survived by all
   but three passengers should not have caused an abrupt stop in the
   flight data record.  Again I cannot recall the Airbus reply.

3: The final seconds of the record showed forward acceleration.  The
   airbus Chief Eng claimed that Davis had this graph upside down and
   a positive reading indicated deceleration.  He also claimed that
   the deceleration was caused by the trees, and that Davis was
   incompetent if he did not know that this format was an
   international standard.  Equinox stated that the international
   standard was for a forward acceleration to give a positive reading
   and that this was the one used by the A320.  Airbus later stated
   that the CE had been referring to a French standard.

4: The final seconds of the record also showed the pilot giving full
   stick back but being overridden by the computer.  The CE stated
   that this was the safety systems stopping the aircraft from
   stalling.  Equinox said something about the Pilot manuals saying
   that at the indicated airspeed the aircraft should have been able
   to climb.

The possibility of an engine compressor stall leading to loss of power was
discussed.  According to Equinox this would lead to a small explosion (I assume
this would be as unburnt fuel vapour was pushed out of the tailpipe) and a drop
in power.  A survivor and a ground witness stated that they had heard such
explosions, but Airbus deny they occurred and point out that no such explosions
are audible on the videotapes.  An early transcript of the CVR did include the
text "(boume) (boume)" (sp?).  Airbus claim this is the sound of impact with
trees.

About 30 seconds were devoted to a pilot employed by Airbus who had publicly
spoken out in support of Capt. Asseline (sp?) who was the pilot at Habsheim.
This pilot claimed that 4 days later he was given an unscheduled medical
examination and had his license withdrawn due to "mental instability".

The authenticity of the "black boxes" recovered from the crash was
questioned.  Officially the boxes are being held by a French court.
Equinox was not allowed to film these, but a magistrate looked at a
video alleged to be of the boxes immediately after being removed from
the crash site and stated that if these were in fact the boxes from
the A320 then something was very wrong.  The implication was that the
boxes delivered to the court were not the boxes recovered from the
crash.

A video of the programme can be obtained by phoning +44 532 438283
ext. 4060 or 4075.

BTW, one of the interviewees had a box file labeled "RISKS" in the
background.  Perhaps he could fill in the holes in my report.  Thanks.

Please note that this report is in no way connected with my employers.  Paul.


Novel on corporate computer espionage

<pbrewer@urbana.mcd.mot.com>
Tue, 02 Oct 90 10:07:13 CDT
Corporate espionage by computer is the subject of a new novel _The
Fool's Run_ by John Camp.  When plans for the latest fighter plane
target acquisition hardware and software are stolen, a defense
contractor decides that only by sabotaging the development work of a
competitor can it be sure of being the only company in a position to
demonstrate the system by the deadline.  The company hires Mr. Kidd
(artist, software designer, former commando) to invade the competitor's
computers and disrupt their operations for a few weeks.  They say:

    the best way ... is through their computer systems--design systems,
    accounting systems, information systems, scheduling and materials.
    Altering them, destroying them, faking them out.

In the style of a classic caper novel, Kidd assembles a team including a
burglar and a sleezy reporter and attacks the defense contractor,
disrupting their operations from all sides.

The author handles the computer entry techniques well.  There is only a
small amount of "magic" involved, and most of that is performed in the
background by "Bobby" (a former phone-phreak we meet only by way of a
data link) who handles such things as telephone trace bypasses.  The
discussions of computer security techniques are right on target, and the
supposed level of security at the target company is on par with what
I've seen at several of the places I've worked.  When it comes to the
actual disruptions things get a little fuzzier, although not to the
point that it fails to work as a novel.

In real life, most malicious computer attacks have been committed by
disgruntled employees or former employees.  Most computer viruses have
been written by misguided enthusiasts.  I haven't heard of this kind of
attack against one company by another.  That doesn't mean it hasn't
happened, and it certainly doesn't mean that it won't happen.  I fear,
this book may give some people ideas.

Camp, John _The Fool's Run_ ISBN 0-451-16712-0 Signet $4.95

Philip Brewer                   pbrewer@urbana.mcd.mot.com
Motorola Urbana Design Center   ...!uiucuxc!udc!pbrewer


CERT Advisory - NeXT systems

<cert-advisory-request@cert.sei.cmu.edu>
Tue, 2 Oct 90 14:57:03 -0400
CA-90:06                       CERT Advisory
                  October 2, 1990
                          NeXT's System Software

This message is to alert administrators of NeXT Computers of four
potentially serious security problems.

The information contained in this message has been provided by David Besemer,
NeXT Computer, Inc.  The following describes the four security problems,
NeXT's recommended solutions and the known system impact.

  Problem #1 DESCRIPTION:  On Release 1.0 and 1.0a a script exists in
  /usr/etc/restore0.9 that is a setuid shell script.  The existence of
  this script is a potential security problem.

  Problem #1 IMPACT:  The script is only needed during the installation
  process and isn't needed for normal usage.  It is possible for any
  logged in user to gain root access.

  Problem #1 SOLUTION:  NeXT owners running Release 1.0 or 1.0a should
  remove /usr/etc/restore0.9 from all disks.  This file is installed by
  the "BuildDisk" application, so it should be removed from all systems
  built with the standard release disk, as well as from the standard
  release disk itself (which will prevent the file from being installed
  on systems built with the standard release disk in the future).  You
  must be root to remove this script, and the command that will remove
  the script is the following:

  # /bin/rm /usr/etc/restore0.9

                                    ---

  Problem #2 DESCRIPTION:  On NeXT computers running Release 1.0 or
  1.0a that also have publicly accessible printers, users can gain
  extra permissions via a combination of bugs.

  Problem #2 IMPACT:  Computer intruders are able to exploit this security
  problem to gain access to the system.  Intruders, local users and remote
  users are able to gain root access.

  Problem #2 SOLUTION:  NeXT computer owners running Release 1.0 or
  1.0a should do two things to fix a potential security problem.
  First, the binary /usr/lib/NextPrinter/npd must be replaced with a
  more secure version.  This more secure version of npd is available
  through your NeXT support center.  Upon receiving a copy of the more
  secure npd, you must become root and install it in place of the old
  one in /usr/lib/NextPrinter/npd.  The new npd binary needs to be
  installed with the same permission bits (6755) and owner (root) as
  the old npd binary.  The commands to install the new npd binary are
  the following:

  # /bin/mv /usr/lib/NextPrinter/npd /usr/lib/NextPrinter/npd.old
  # /bin/mv newnpd /usr/lib/NextPrinter/npd
      (In the above command, "newnpd" is the npd binary
      that you obtained from your NeXT support center.)
  # /etc/chown root /usr/lib/NextPrinter/npd
  # /etc/chmod 6755 /usr/lib/NextPrinter/npd

  The second half of the fix to this potential problem is to change the
  permissions of directories on the system that are currently owned and
  able to be written by group "wheel".  The command that will remove
  write permission for directories owned and writable by group "wheel"
  is below.  This command is all one line, and should be run as root.

  # find / -group wheel ! -type l -perm -20 ! -perm -2 -ls -exec chmod
  g-w {} \; -o -fstype nfs -prune

                                    ---

  Problem #3 DESCRIPTION:  On NeXT computers running any release of the
  system software,  public access to the window server may be a
  potential security problem.

  The default in Release 1.0 or 1.0a is correctly set so that public access
  to the window server is not available.  It is possible, when upgrading from
  a prior release, that the old configuration files will be reused.  These
  old configuration files could possibly enable public access to the window
  server.

  Problem #3 IMPACT:  This security problem will enable an intruder to gain
  access to the system.

  Problem #3 SOLUTION:  If public access isn't needed, it should be disabled.

  1. Launch the Preferences application, which is located in /NextApps
  2. Select the UNIX panel by pressing the button with the UNIX
     certificate on it.
  3. If the box next to Public Window Server contains a check, click on
     the box to remove the check.

                                    ---

  Problem #4 DESCRIPTION: On NeXT computers running any release of the
  system software, the "BuildDisk" application is executable by all users.

  Problem #4 IMPACT:  Allows a user to gain root access.

  Problem #4 SOLUTION: Change the permissions on the "BuildDisk" application
  allowing only root to execute it.  This can be accomplished with the
  command:

  # chmod 4700 /NextApps/BuildDisk

  To remove "BuildDisk" from the default icon dock for new users, do the
  following:

  1. Create a new user account using the UserManager application.
  2. Log into the machine as that new user.
  3. Remove the BuildDisk application from the Application Dock by dragging
     it out.
  4. Log out of the new account and log back in as root.
  5. Copy the file in ~newuser/.NeXT/.dock to /usr/template/user/.NeXT/.dock
    (where ~newuser is the home directory of the new user account)
  6. Set the protections appropriately using the following command:
        # chmod 555 /usr/template/user/.NeXT/.dock
  7. If you wish, with UserManager, remove the user account that you created
     in step 1.

  In release 2.0, the BuildDisk application will prompt for the root password
  if it is run by a normal user.

 = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =

CONTACT INFORMATION

For further questions, please contact your NeXT support center.

NeXT has also reported that these potential problems have been fixed in
NeXT's Release 2.0, which will be available in November, 1990.

Thanks to Corey Satten and Scott Dickson for discovering, documenting, and
helping resolve these problems.

Edward DeHart, Computer Emergency Response Team/Coordination Center (CERT/CC)
Software Engineering Institute, Carnegie Mellon University Pittsburgh, PA
15213-3890       E-mail: cert@cert.sei.cmu.edu
Telephone: 412-268-7090 24-hour hotline: CERT personnel answer
           7:30a.m.-6:00p.m. EST, on call for emergencies other hours.

Past advisories and other information are available for anonymous ftp
from cert.sei.cmu.edu (128.237.253.5).


Fair Information Principles

Jeff Johnson <jjohnson@hpljaj.hpl.hp.com>
Wed, 03 Oct 90 14:40:57 PDT
This is a summary of the Fair Information Principles, excerpted with permission
from an e-mail message sent by Marc Rotenberg, directory of CPSR's Washington
D.C. office and head of its Computers and Civil Liberties Project.

I thought RISKS readers might be interested.

FAIR INFORMATION PRINCIPLES

The Fair Information Principles were developed by a U.S. Government
Study Committee in 1973, chaired by Willis Ware of the Rand
Corporation.  Shortly after the commission released its final report
("Records, Computers and the Rights of Citizens"), Congress passed
comprehensive privacy legislation:  the Privacy Act of 1974.  Much of
the privacy law that followed the Privacy Act (e.g., the Right to
Financial Privacy Act, the Cable Policy Act, the Electronic
Communications Privacy Act, and the Video Privacy Protection Act) are
based on the Fair Information Principles.

Many other countries follow the Fair Information Principles.  Recently
in Paris, European Data Protection commissioners recommended that the
EC 92 charter include mandatory provisions for the enforcement of Fair
Information Principles across all European countries.

These are the Principles:

1. There must be  a way for a person to prevent information about the
person that was obtained for one purpose from being used or made
available for other purposes without the person's consent.

2. There must be no personal data record-keeping systems whose very
existence is secret.

3. There must be a way for a person to find out what information about
the person is in a record and how it is used.

4. There must be a way for a person to correct or amend a record of
identifiable information about the person.

5. Any organization creating, maintaining, using, or disseminating records of
identifiable personal data must assure the reliability of the data for their
intended use and must take precaution to prevent misuses of the data.

Jeff Johnson, HP Labs


Television rating (nee universal listening) device

Tim Wood at home <tim@axolotl.UUCP>
Mon, 1 Oct 90 11:30:28 PDT
Found in the Oakland Tribune "Patents" column, Oct. 1, reprinted from
The New York Times:

        In search of a more accurate way to measure television and
    radio audiences, a small company in Chicago has patented a
    pocket-sized device that silently monitors and logs the
    programs a person listens to.
        The battery-powered device is based on "acoustic matching."
    [this term is not precisely defined] ... [A] microphone senses
    sounds near the person being monitored and a microprocessor
    converts these sounds into a digital code.
        ... Users would place the monitoring devices on
    battery chargers when they go to bed.  The battery charger
    would be connected to a telephone line, enabling the device to
    transmit the day's data to a central computer at the audience
    measurement company.

Hope all of your RISKS alarms are ringing as loudly as mine are.  The
frightening prospect of creation of libraries of users' private sounds
comes to mind.  As does the funny, if Machiavellian, image of public
broadcasting of these sounds, a la the tryst between Majors Hoolihan
and Burns in the movie "MASH."

This development is interesting in light of (what I see as) a duality
in society`s view of high tech of simultaneous infatuation and distrust.
Hopefully the latter view will be applied to the new device.
-TW


From under a Rock???

Ed Hall <edhall@rand.org>
Thu, 04 Oct 90 11:56:40 PDT
I've been loosely following the various ``subliminal message'' lawsuits which
have been winding their way through the courts recently.  These are
product-liability suits alleging that subliminal messages in rock music have
driven people to suicide.  One such case, against British group Judas Priest,
was recently dismissed.  Another against Ozzy Osbourne is now pending.

There is a computer RISK here.  According to today's Los Angeles Times:

    ...  Sound Analyst Evans [a lecturer at Univ. of Nevada with
    masters degrees in physics and computer science] said she had
    spent about a month analyzing audio subliminal messages
    allegedly implanted on the "Blizzard of Oz" cassette using the
    same home-computer software package employed in the Judas Priest
    case. ...

I can only guess at what this "home-computer software package" is. (If
anyone has additional information about it, please let me know).  One
thing I'm sure of, however: it hardly affords an accurate model of human
auditory perception (unless its author has managed to leapfrog what
would no doubt be decades of neurophysiological research).  Its use in
court no doubt arises from the persisting association of The Computer
with unchallengeable accuracy and authority.

I foresee nothing but trouble in the interaction between the notion of
"subliminal messages" (whether auditory or visual) and the increasing
capability for computers to perform extensive signal processing--whether that
"processing" is meaningful or not.  As the recent "Face on Mars" flap
illustrates, people will see (or hear) just what they want to see (or hear),
given the tools to create "evidence".  Computers greatly enhance the power for
self-delusion.
                    -Ed Hall,       edhall@rand.org
[Disclaimer: This all is my personal opinion ONLY.]

Please report problems with the web pages to the maintainer

Top