The RISKS Digest
Volume 16 Issue 67

Friday, 23rd December 1994

Forum on Risks to the Public in Computers and Related Systems

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

Please try the URL privacy information feature enabled by clicking the flashlight icon above. This will reveal two icons after each link the body of the digest. The shield takes you to a breakdown of Terms of Service for the site - however only a small number of sites are covered at the moment. The flashlight take you to an analysis of the various trackers etc. that the linked site delivers. Please let the website maintainer know if you find this useful or not. As a RISKS reader, you will probably not be surprised by what is revealed…


Software glitch snares Social Security Administration
Mike Manos
Cancelbot Derails Online Promo
WSJ via Edupage
Ben & Jerry's expects first loss
Arthur D. Flatau
Proliferation of Cockpit Warning Signals
David Walter
Year 2000 date problems already happening now
Re: Prevention of Oral Hacking
Brad G. Parks
Re: Pentium FDIV problem - so what's new?
Mark Brader
Intelligent Commentator on McNeil-Lehrer
D.P. Schneider
Financial Payment Systems
Jeff Stapleton
Possible solution to the (electronic) meme problem
Bob Mehlman
Advertising on the Net
Mich Kabay
Risk Takers Among Management
Tom Kaiser
Re: Koppel et al.
Peter da Silva
Re: Microsoft and the Catholic Church
John Stevenson
Timothy Hunt
Public CM WAIS Server to end operation
WAIS Admin
Call for Papers and Panels: National Information Systems Security Conf.
Jack Holleran
Info on RISKS (comp.risks)

Software glitch snares the Social Security Administration

Mike Manos <>
Wed, 21 Dec 1994 14:49:13 EST
[from Federal Computer Week, 21 Nov 1994]

The Social Security Administration has uncovered a software coding error
that over the years nickeled and dimed Social Security recipients out of
$478.5 million.  The problem occurred in 1978, when SSA wrote new code
because employers began reporting earnings annually rather than quarterly.
The software program, with its 76 sets of variables, had a few holes in it.
Over 16 years, about 426,000 people did not receive appropriate payment

Cancelbot Derails Online Promo (WSJ via Edupage 12/20/94)

Edupage <>
Tue, 20 Dec 1994 18:15:26 -0500
Messages promoting "Netchat," a new book by Michael Wolff, have been
unceremoniously wiped out by someone calling him/herself "Cancelmoose." The
vigilante, who used a program called a cancelbot, worked through a computer
in Finland that enables a mail sender to remain anonymous, and claims to
have performed similar services more than a dozen times in the past.
(*Wall Street Journal*, 20 Dec 1994, p. B7)

Ben & Jerry's expects first loss

Arthur D. Flatau <>
Wed, 21 Dec 94 08:52:14 CST
According to an article (page F2) in the Tuesday December 20, 1994 issue of
the Austin American Statesman (from the Associated Press) Ben & Jerry's
Homemade Inc. is going to report its first quarterly loss since it went
public.  Quoting some of the relevant to risks part of the article:

        Adding to the bad news, the company said Monday that recurring
        problems in a comuterized handling system will force it to
        delay opening a new, more efficient manufacturing plant in
        St. Albans, Vt. until the middle of next year.  [...]

        "While out earnings have been negatively impacted in 1994 by
        the inefficiencies in the company's production planning,
        purchasing and inventory management systems ... the
        anticipated loss in the current quarter is due primarily to
        lower than expected sales lever," [Ben & Jerry's President
        Chuck] Lacy said in a written statement.  [...]

        At the new St. Alban's plant, a computerized system that was
        supposed to automate handling of ice cream after it has been
        packed into pints has software problems, forcing the company
        to substitute a more conventional system.
    That likely will require installation of different equipment
        and more employees for handling, although the company said it
        will still be more efficient than current manufacturing.

    The St. Albans plant is designed to consolidate all of Ben &
        Jerry's manufacturing in Vermont again.  The company hires
        Edy's [a major competitor] of Fort Wayne, Ind. to make about
        37 percent of its ice cream, [Ben & Jerry's spokesman Rob]
        Michalak said.

Art  Computational Logic, Inc.  Austin, Texas

Proliferation of Cockpit Warning Signals

Wed, 21 Dec 94 13:54:23 GMT
The avionics thread seems to have gone quiet, perhaps because none of the
several recent airliner crashes has involved an A320 aircraft. Here is an
account of an incident I came across recently which highlights the problem
of proliferation of warning signals in complex human-machine interfaces.

The incident involved a Continental Airlines 727 which came perilously
close to landing gear-up at Chicago O'Hare about a year ago.

Initial investigation indicated that the crew had become distracted
from their landing check-list by TCAS (Traffic Collision Avoidance
System) warnings which sounded repeatedly as they flew through the busy
airspace around O'Hare. The 727's gear warning did not trigger because
the flap setting was 25 degrees instead of the usual 27.5 (in order to
maintain speed on approach as requested by air traffic control). A ground
proximity warning did sound at 500 ft, but lack of gear is only one of
a number of possible causes for this warning.

A taxiing American Airlines pilot saw the 727 on short final with no gear.
The tower radio transcript makes hair-raising reading:

    American:   "Tower, Continental's got no gear!"

    Tower:      "Who's got no gear?"

    American:   "Continental, just about ready to touch down
            on two-seven left"

    The ground controller shouted to his colleague who was controlling
    Continental's approach, who warned:

            "Go around, go around Continental. No gear!"

The 727 succeeded in executing a go-around, but not before scraping the rear
third of the fuselage on the runway. It landed normally out of the second

David Walter

Year 2000 date problems already happening now

Elana who? <>
Fri, 23 Dec 1994 05:48:33 GMT
A few months ago, there was an excellent article in RISKS about the problems
of the year date rolling from xx/xx/99 to xx/xx/00 and how it could affect
computer systems. Today I mentioned the article in passing to a rep at
my local bank as I was making a routine deposit.  She said:  "It's happening
already."  Apparently there is a problem with some of the credit cards
they have been issuing to their customers.  The expiration date on them
is the year 2000, and for some tech reason she could not explain, two
major stores here in town refuse to take them "because that weird date
messes up their computers."

I regret that I cannot offer more info than this, but it is serious food
for thought...


   [See the latest American Scientist, Jan-Feb 1995, which has a very
   nice article by Brian Hayes, spanning the entire subject.  (Much of
   it is drawn from the RISKS Archives.)  PGN]

Re: Prevention of Oral Hacking

Brad G. Parks <>
Thu, 22 Dec 94 11:21 CST
   I am a big fan of Voice Recognition Technology, however trite
and useless people may say it is.  And one thing that people have
failed to mention so far that there are ways to prevent/discourage
oral hacking (at least with Macintosh's general program).
   With the Mac program there is an option that allows the user to
name the computer something other than "Computer."  There is another
option for setting an "Attention Key" to basically give the program
earplugs until that key is pressed again.  There is also a voice
recognizing option that sets how precise your language must be.
   It is doubtful, at least to me, that somebody running around the
office yelling "Computer, Restart, Yes" would restart my Mac, whose
name is Alfred, has the precision level set rather high, and also
has the voice recognition asleep until i press Keypad-0.

Brad G. Parks (    [Software Engineer, O-Programmer]
Introl Corporation (612)-631-7810    2675 Patton Rd. St.Paul, MN  55113

  [Ah, now we know!  And we can record your normal voice and replay it. PGN]

Re: Pentium FDIV problem - so what's new? (Peterson, Risks-16.66)

Wed, 21 Dec 1994 14:03:47 -0500
It occurs to me that the English vocabulary might be related to why
people have reacted so strongly to the existence of this bug — which,
let's face it, is not all that devastating *relative to other bugs*.
The term "computer" suggests that computation is the most important
and fundamental thing that the machine does, and because of this, any
bug directly related to computation is seen as more serious than others.

In French the machine is called an "ordinateur", which is related to our
word "ordinal" and conveys connotations of *organizing* data.  What we
mostly call computer science, they call "informatique" — the study of
information.  And if you think about what computers are used for today,
you will see that these terms are at least as fitting as the English ones.
In English when we really mean that computation is the important part of
the work, we use terms like "numerical" in addition to "computing".

Would English speakers have reacted equally strongly to a — no pun intended
-- comparable bug in the instructions for comparing and branching?  I think
they would not.

(No, you don't need to tell me that a large part of the reaction in this
case was to Intel's reaction, not to the original problem.)

Mark Brader  SoftQuad Inc., Toronto

Intelligent Commentator on McNeil-Lehrer

Thu, 22 Dec 94 16:34:00 PST
A recent McNeil-Lehrer report (I'm pretty sure it was Tuesday night's  --
12/20/1994) had an intelligent commentator as an interviewee on the Pentium

I agree with his comments about the Pentium (that the problem was probably
more severe than Intel admitted and less severe than IBM trumpetted), but
the reason I'm pointing this interview out is because Robin McNeil asked
something to the effect, "So the issue is also that we shouldn't trust an
answer just because the computer gave it to us?"  The interviewee answered
that this was just so, and that we should keep asking for and demanding
proof that the computer is giving good answers.

There were also what I thought were intelligent comments on the difficulty
of thoroughly testing such complex parts, and that testing had to involve
strategies to maximize the confidence obtained.

I hope that one of your other submitters has given (or will give) the
identity of the interviewee, and some exact quotes.  I was impressed by his
grasp of the big picture, and that he managed to convey it to the
interviewer, too.  (I think the choice of interviewer helped, but even
intelligent interviewers can't be prepared to understand all issues, and
McN-L tend to have more news on political and economic issues than on
technical issues).

Hope you caught the interview, too!


Financial Payment Systems

Thu, 22 Dec 1994 10:56:25 -0600
In regards to RISKS-16.64 here are some thoughts and concerns about
financial payments on the World Wide Web, CommerceNet, the Internet, and
electronic media in general:

Debit card payments today require an additional piece of authenticating data
- the Personnel Identification Number or PIN. There are ISO (9564) and ANSI
(X9.8) standards which tell us how to securely formulate a PIN block and
encrypt it using the Data Encryption Algorithm or DEA (X3.92).  Credit card
payments may soon also require the PIN.  However, we have no standards today
telling us how to encrypt PINs using public key algorithms.  True, the
advent of the smartcard may eliminate the transmission of PINs, but not for
a while yet (years?).

Speaking of public key cryptosystems, there ARE existing ISO and ANSI
standards available today:
     ISO-CD 11568 parts 4 and 5 just completed the public ballot process.
     ANSI X9.30 part 1 - The Digital Signature Algorithm (DSA)
     ANSI X9.30 part 2 - The Secure Hash Algorithm (SHA)
     ANSI X9.30 part 3 - Certificate Management for the DSA is completing its
                         public ballot process.

However, due to the legal problems between Cylink Corporation and RSA Data
Security on fair licensing agreements for the RSA patents, the matching set
of ANSI standards (X9.31 parts 1, 2, and 3) for RSA may be put on hold by
the Accredited Standards Committee (ASC) X9 - for further details, call:
     X9 secretariat, American Bankers Association (ABA), 202-663-5284
     X9F chairman, Marty Ferris, 202-622-1110
     X9F1 chairman, Blake Greenlee, 203-762-8580

Although the standard addresses the DSA, X9.30-3 Certificate Management is
rather generic and deals with the roles and responsibilities of the
asymmetric key pair owners, the certificate authority(s), and the public key
users.  Interesting stuff and can (and should be) used for implementing any
public key cryptosystem.  Copies can be purchased from the ABA (plug).

Jeff J. Stapleton - MasterCard International

Possible solution to the (electronic) meme problem

Wed, 21 Dec 1994 18:43:08 PST
Keith Henson (RISKS 14.63) points out that computer viruses and human fads
fit Richard Dawkins' concept of a meme, adding

  What computers and fast communications have done is to radically change the
  replication constants for memes, and to make unworkable the controls society
  has tried to put on such things.

In "Mind Viruses" (American Scientist, Jan-Feb 1995, p.26), Michael Szpir
discusses 'human information parasites', such as the St. Jude 1 chain letter,
likened to memes by Dawkins and Oliver Goodenough (Nature, Sept 1, 1994).

Some years ago, I received an electronic chain letter, complete with lengthy
sequential history of previous senders.  These letters differ from the old-
fashioned kind — they always have return addresses.  It occurred to me that
the appropriate response was to SEND IT BACK, and encourage the sender and all
others to do likewise, hoping an avalanche of returned mail would eventually
swamp the original sender, if it didn't bring the network to its knees.

No such luck.  I actually tried it, forwarding the letter back to the
sender with subject LET'S SEND IT BACK.  Unfortunately, he returned it
immediately, longer by two headers, with subject THAT'S NOT FAIR...IT'S
GOT TO BE 5 OTHER PEOPLE.  I did not repeat the experiment, though I tried
unsuccessfully to get others to do so.

Still, I wonder, next time a chain letter or other intentional meme spreads
through the Internet, if everyone returned their copies to the sender ...

Bob Mehlman, UCLA/IGPP

Advertising on the Net

"Mich Kabay [NCSA Sys_Op]" <>
22 Dec 94 14:25:04 EST
In the 05 Dec 1994 issue of _Network World_ on p. 41, "John Doe" recounts
his tale of woe after advertising on the Internet ("Ad leads to a nightmare
on Cyber-Street).  In brief, his main points are

* Placed ads in several USENET groups.
* Swamped with email abuse and flames in the groups as a result.
* Someone posted his 800- number in groups as a phone-sex #.
* Switchboard swamped with requests for phone sex; operators horrified.
* One switchboard op quit because of what callers were saying.
* All 800- calls switched to the originator of the advertising as punishment.
* Impractical to cancel 800- service because of special letter combinations.

Moral:  don't advertise on the Net.

<<Comments by MK:

One issue is whether it's good or bad to advertise on the Net.

Another is whether it's good or bad to respond to advertising on the Net in the
ways demonstrated in this case.

Yet another is that exactly the same techniques could be used against any
organization even if it _doesn't_ advertise on the Net.<>

M.E.Kabay,Ph.D./DirEd/Natl Computer Security Assn

Risk Takers Among Management

Tom Kaiser <>
Thu, 22 Dec 1994 23:22:43 -0500
Subject: Upper Managements View of Information Security(Infoworld 12/19/94)

The following extract from _INFOWORLD_ 12/19/94 demonstrates some of the
risks inherent in human management. (I read this shortly before reading
RISKS 16.66 and the posts calling for papers and serious work in all areas
of network and computer security).

42% of senior management at U.S. companies consider information security to
be only "somewhat important" or "not important," according to a survey by
Ernst and Young LLP, with 15% of companies devoting no full time resources
to security. This though even more than half the respondents reported
information-related losses.  Also , among respondents who said they were
running "mission-critical applications" on their LAN's, half reported that
security was unsatisfactory. What will it take to make security a priority
in industry? The Stealth bomber blueprint on place mats in Pizza Hut?  (end

We can only hope that some of the recent antics and activities on the
network cause these top managers to focus some their energies, and resources
on the protection of their (our?) data.  Analysis of all probable problems
and or threats should precede implementation of any changes to or
introduction any systems.  I think most readers of RISKS would enjoy reading
the entire poll and possibly having helped in phrasing some of those obscure
questions, and "open to interpretation answers."

Thomas Kaiser

Re: Koppel et al. (RISKS-16.66)

Peter da Silva <>
Thu, 22 Dec 1994 17:51:54 -0600
I seem to recall someone with the name of "Hughes" being forced to change
the name of his hotel because Howard Hughes objected. Be glad your name
doesn't match someone like that.

Re: Microsoft and the Catholic Church (RISKS-16.66)

John Stevenson <>
Fri, 23 Dec 1994 22:53:37 +0000
>Microsoft has no plans to acquire Catholic Church (from EDUPAGE)

Surely the big RISK here is that the world is infested with people who lack
even a trace of a sense of humour. Hoax? I believe 'Microsoft to acquire
Catholic Church' was what we writers call 'a joke'.

John Stevenson

   [I would call it a *parody*.  RISKS readers seem to be a cut above the
   average in terms of not being fooled so easily, especially by such obvious
   parodies.  But remember there are a lot of nontech folks out there with
   very bad spoof-suspicion thresholds.  Even the most *obvious* humor
   sometimes gets taken seriously.  Remember what happened to Piet
   Beertema's Chernenko hoax!  Many people took it very seriously.  Some
   might have even attempted to spoof the spoofer by pretending to take it
   seriously.  At any rate, best wishes for a happy new year, with its own
   collection of good humor.  PGN]


"Timothy Hunt [Assistant Unix Systems Manager]" <>
Wed, 21 Dec 94 14:55:07 +0000
Noticing the article that made reference to EDUPAGE [Microsoft has no plans to
acquire Catholic Church], RISKS readers may be interested to know that EDUPAGE
is available via the Web, with URL :

Timothy J. Hunt, Joint Dept. of Physics, The Inst. of Cancer Research, Downs
Road, Sutton, Surrey England SM2 5PT +44(0)181 642 6011 x3312

Public CM WAIS Server to end operation

Wais Administrator account <>
Thu, 22 Dec 94 11:36:41 EST
The Public CM WAIS Server, which has provided a searchable database of past
RISKS traffic as risks-digest.src, will cease operation on 27 December 1994.
Please update the "Info on RISKS" item and inform your readers of the change.

  [Here we come a-WAIS-ailing!  And many thanks for your past serv-ice!  PGN]

Call for Papers and Panels: National Information Systems Security

Jack Holleran <Holleran@DOCKMASTER.NCSC.MIL>
Thu, 22 Dec 94 20:13 EST
          (formerly the National Computer Security Conference)

Co-sponsored by the National Computer Security Center and
the National Institute of Standards and Technology

Baltimore Convention Center, Baltimore MD
October 10-13, 1995

The National Information Systems Security Conference audience represents a
broad range of information security interests spanning government, industry,
commercial, and academic communities.  Papers and panel discussions typically

 (  research and development for secure products and systems presenting
      the latest thinking and directions;
 (  practical solutions for real-world information security concerns;
 (  implementation, accreditation, and operation of secure systems
      in a real-world environment;
 (  evaluation of products, systems, and solutions against trust criteria;
 (  security issues dealing with rapidly changing information technologies;
 (  network security issues and solutions;
 (  management activities to promote security in IT systems including
      security planning, risk management, and awareness and training;
 (  international harmonization of security criteria and evaluation;
 (  social and legal issues such as privacy, ethics, investigations,
      and enforcement;
 (  tutorials on security basics and advanced issues; and
 (  highlights from other security forums.

We invite the submission of papers and proposals for panel discussions in any
of the above areas as well as other topics related to IT system security.  We
especially encourage student papers written by individuals in degree programs.
The student should not have been previously published, and the paper shall be
endorsed by an academic advisor.

BY MARCH 1, 1995: Eight (8) copies of your draft should arrive at the
following address.  See instruction on reverse side regarding the format of
your submission and accompanying information required.
                    National Information Systems Security Conference
                    Attn: Conference Secretary, APS XI
                    National Computer Security Center
                    Fort George G. Meade, MD 20755-6000

BY JUNE 1, 1995: Authors and Panel chairs selected to participate in the
Conference will be notified and advised when final papers and panel statements
are due.

For additional information on submissions, please call (410) 850-0272 or send
Internet messages to NISS_Conference@DOCKMASTER.NCSC.MIL.  For other
information about the National Information Systems Security Conference, please
call (301) 975-2775.


Cover Sheet:  Type of submission (paper, panel, tutorial)
              Title or Topic
              Abstract (not to exceed 250 words)
              Organizational Affiliation
              Phone numbers (voice and fax, if available)
              Internet address, if available
              Point of Contact, if more than one author

Submissions related to work under U.S.  Government sponsorship must also
include the following information:

              U.S. Government Program Sponsor or Procuring Element
              Contract Number (if applicable)
              U.S. Government Publication Release Authority

Classified material or topics must NOT be submitted.

Draft Papers:
              10 page maximum, including figures and references.
              Include title, abstract, and keywords on first page.
              No more than 12 characters/inch and 6 inches/line.
              One-inch margins all around.

Since the paper referee process will be anonymous, names and affiliations of
authors should appear only on the separate cover sheet.  All submissions are
treated as proprietary information belonging to the authors.

Release for Publication and Copyright: Authors are responsible for obtaining
government or corporate releases for publication.  Written release will be
required for all papers to be published.  Papers developed as part of official
U.S.  Government duties may not be subject to copyright.  Papers that are
subject to copyright must be accompanied by written assignment for multi-media
publication to the National Information Systems Security Conference Committee.

 (  Panels are limited to 90 minutes, including time for prepared
      remarks, discussion, and audience interaction.
 (  Proposals are not to exceed two pages but should summarize the topic,
      issues, viewpoints and questions that will be addressed by the panel.
 (  Proposals are to include the names of panelists, panel chair, and
      affiliation of each participant.
 (  Each panel is limited to *five (5)* persons, including panel chair
      and four (4) panelists.
 (  Panels will be selected by the Conference Committee.  Panel chairs
      and panelists will be expected to provide written statements for
      inclusion in the Conference Proceedings.
 (  Panel proposals *must be* received by March 1, 1995.

Please report problems with the web pages to the maintainer