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 21 Issue 56

Thursday 2 August 2001


NASA data from 1970s lost due to "forgotten" file format
Aaron Dickey
Motorola Stock Drops 99.95%!
Daniel Norton
JDS Uniphase quarterly results hacked? NO!
Dave Isaacs
Freeware app to retrieve passwords from Internet Explorer
Lyle H. Gray
Totally Hip with spyware
Michael F. Maggard
Medical records via e-mail
William Colburn
AS IF: draft-ietf-dnsext-ad-is-secure-03.txt
John Gilmore
Microsoft's PGP keys don't verify
Brian McWilliams
Telling all to the police
Norm deCarteret
Identity theft
Jack Holleran
Risks of profanity filtering
Paul Bissex
Car-door lock remote control activates another car's alarm
Mark Brader
Mike Albaugh
Re: MSN security upgrade forces new e-mail address
Robert J. Woodhead
No Appleplexy needed
Dave Stringer-Calvert
Re: Autoresponder goes haywire
Richard Johnson
Re: Erroneous air navigation reading
Mike James
Re: Polarized sunglasses and LCD frustration
Chris J Dixon
Info on RISKS (comp.risks)

NASA data from 1970s lost due to "forgotten" file format

<Aaron Dickey <>>
Sat, 28 Jul 2001 23:31:43 -0700 (PDT)

In 1999, USC neurobiologist Joseph Miller asked NASA to check some old data
the Viking probes had sent back from Mars in the mid-1970s. Miller wanted to
find out whether certain information on gas released by Martian soil, which
at the time had been dismissed as meaningless "chemical activity," was
actually evidence of microbial life. NASA found the tapes he requested, but
they didn't find any way to read them. It turns out that the data, despite
being only about 25 years old, was in a format NASA had long since forgotten
about. Or, as Miller puts it, "The programmers who knew it had died."

Luckily, Miller has been able to cobble together about a third of the data
and get some useful results, but only because some form of printed record
had been saved. (And yes, he does believe the Viking probes turned up
evidence of microbes.)

Source: Reuters. Original article is available, at least temporarily, at
<>, or

Motorola Stock Drops 99.95%!

<Daniel Norton <>>
Thu, 02 Aug 2001 10:17:06 -0400

My Yahoo! alerts window popped up with an explosive sound this morning to
notify me that Motorola's stock (MOT NYSE) value crashed.  Incredulous, I
went to the Yahoo! Finance page and confirmed it.  Undaunted, I proceeded to
the NY Times finance site which only concurred.  Finally the NYSE site
confirmed that, in fact, the value of MOT had been exactly one penny
(US$0.01) at the open, but rebounded spectacularly, even to exceed the
previous day's close!  (cf. )

Hopefully, anyone who automates trading programs around this kind of
glitch (It _was_ a glitch, wasn't it?), but we RISKs readers know that
such hopes aren't always fulfilled.

Daniel Norton

JDS Uniphase quarterly results hacked? NO!

<Dave Isaacs <>>
Mon, 30 Jul 2001 09:39:49 -0400

I saw this interesting aside in an *Ottawa Citizen* article (27 Jul 2001)
about JDS Uniphase's latest quarterly results:

"The world's largest maker of fibre optic components was forced to halt the
trading of its stock for most of the afternoon yesterday because a hacker
broke into its corporate network and stole a draft copy of the company's
fourth-quarter results. It had been released before the markets closed
yesterday afternoon."

The article is at

The obvious risk here is the consequences of storing very valuable
information unencrypted on a network-accessible computer.  Nothing new in
that lesson.  What would be interesting is knowing is *how* JDS Uniphase
knew that this break-in had occurred, and what form the break-in took.  It
sounded like a story we'd all be interested in hearing.

A further article, from the *Globe & Mail* (28 Jul 2001), with the rather
convoluted URL of,C/20010
contains more details.  Apparently, there was no 'hacker' or 'break-in'.
JDS had placed the release on their Web site.  A sharp-eyed surfer noticed
that if you type in the exact file name, up pop the results.  I suspect that
a document-naming convention was apparent from looking at previous financial

As to how JSU found out about the 'break-in':  the 'hacker' phoned them up
and told them.

Dave Isaacs <>

  [JDS apparently reported a $51 billion loss for the year ending 30 Jun
  2001, and 16,000 jobs lost.  PGN]

Freeware app to retrieve passwords from Internet Explorer

<"Lyle H. Gray" <>>
Mon, 23 Jul 2001 22:24:15 -0400 (EDT)

The following item appeared in the "Download This" section of the
Earthlink Weekly Email Newsletter on 07/23/2001:

* Windows: Password Recovery
If you tell your browser to save Web site passwords so that you
don't have to reenter them, you might forget those passwords over
time. This program can reveal the passwords hidden behind those
asterisks in Web site login screens. (Freeware)

This item highlights an inherent risk of allowing IE to save your passwords
(other than the obvious one that anyone with physical access to your system
would also have access to your password-protected pages): Someone with
access to your system may be able to determine a pattern to your password
choices (especially if you only have one password...)

  [and also to use your IE passwords directly while masquerading...  PGN]

Totally hip with spyware

<"Michael F. Maggard" <>>
Tue, 31 Jul 2001 17:32:24 -0400

Recently it was discovered that the Mac software "Livestage Pro" by Totally
Hip software has been reporting back its license, usage, and environment to
its manufacturer via a covert http dialogue.

The company has refused to respond to the discovery "officially", but one of
their staff members has been corresponding publicly on the popular Mac
website at There he's expressed
surprise that anyone is concerned and asserts his business has the full
right to include this sort of tracking, that it is noted deep in one of the
readme files and permission to "electronically verify their serial number "
is specified within the software license.

The non-representative goes on to state that in the future Totally Hip
intends to somehow secure the collected information and this is all simply a
legitimate anti-piracy effort. Finally he's taken the Web site to task for
posting letters that detail how to block the reporting function (edit one's
hosts file), likens it to supporting software piracy and closes with
"Honestly we are not an evil conspiring company."

This isn't an isolated incident for Mac software developers; powerhouse
Adobe  has been installing a mysterious file of their own that regularly
"calls home" for reasons unknown. Adobe has promised to explain this new
feature, what it does and what it is communicating but to date have not
followed through.

Medical records via e-mail

<"Schlake ( William Colburn )" <>>
Mon, 30 Jul 2001 15:31:52 -0600

I live in a small town.  Two of the doctors I visit have been my doctors my
entire life, and the third in the office has been a friend of the family
just as long.  The office has been run virtually the same my entire life
(they still have the original monochrome monitors on their office PCs).  The
newest doctor, however, demanded some new-fangled ideas as part of her
contract to work there.  She makes recordings of her medical notes and they
are transcribed by a third party company.  The transcriptions come via
e-mail (which the office can't receive) in MS word 2000 (which the office
can't read), so she gets them at her house and prints them out.  She is out
of town for three weeks now, and they asked me to take care of this for
them.  I don't know if my experience is typical of the medical transcription
business, but I suspect that it is.

The transcriptions are made at the office from the doctor reading her notes
onto tape.  I didn't ask, but I suspect that the tape is then mailed or
shipped to the transcription agency.  The employee there then types up the
information and e-mails it back out to the "appropriate place".  The e-mail
is not (cryptographically) signed, nor is it encrypted.  In the case of my
local office, it goes from the ISP of the person doing the work to a local
ISP here in Socorro.  The doctor has a wireless link from her house to the
ISP and uses an unencrypted POP session.  The transcripts are then launched
from outlook into word, and printed out.  She saves a local copy onto her
hard disk (just in case) and deletes them from the server.

These documents can contain a lot of information.  A medical history will
include tremendous personal data on not just the patient, but on their
entire family, including date of birth and lifestyle.  A simple office visit
can be mundanely bland (a broken wrist) to life-shatteringly personal
(someone here was inspired by "Bobbit"), but always contains the persons
real name, complaint, diagnoses, and prescription.

There are numerous places along the trip that this information could fall
into the wrong hands.  A virus could be present at either end of the e-mail
which might compromise data.  The data passes through ISPs in the clear, and
could be intercepted or modified while in transit.  The wireless ISP is like
a scrolling marquee if someone has the right equipment.  Outlook likes to
"keep e-mail on the server" even if the user has deleted it, so all those
transcripts could still be on the local ISP.  And lastly, a copy of
everything is stored on her computer in her house and not in the security of
the office.

I often tell people that I have "no delusions of privacy" in our modern
world.  It keeps me sane.

AS IF: draft-ietf-dnsext-ad-is-secure-03.txt

<John Gilmore <>>
Sat, 28 Jul 2001 12:16:53 -0700

I think some of you guys have gotten so tied up in micromanaging DNS
Security implementation details that you forgot what swamp we were trying to

There is no point in building a cryptographically-secured DNS in which many
of the machines will be configured to "just believe whatever they are told,
regardless of the cryptographic signatures"!

We already have such a DNS -- today's.  It doesn't need signatures or
AD bits or big packets or any other changes.  Anyone who is happy
with that can go home and stop arguing.  The rest of us are interested
in the real security and integrity of the Internet.

Any client implementation that listens to a single bit of the response to
tell it whether the response is cryptographically valid must be considered
noncompliant with the DNSSEC spec.  It's just an old fashioned insecure DNS
client.  There's nothing wrong with that, as long as you don't have any high
trust expections for it.

Any server which deposits a single bit in the response to claim to clients
that it has cryptographically validated the results, so they don't have to,
is just encouraging the above abuse.

I'm not shocked to find people advocating that such a server actually
lie to the clients about whether it has validated the data.  The
entire model (trusting a packet to tell you whether somebody else has
validated the data) provides ample opportunities for not only your
friends but your enemies to lie to you.  Just like in the current DNS.


PS: I know, I know, the "valid" bit will be secured by some "out of band"
means.  Like a shared static key, and/or by the security of the file system
on the server.  Right.  For extra credit: composing several weak security
primitives produces what?  Strong security or weak security?

PPS: The real question is why anyone is advocating that the DNS be "secured"
by lame security.  There are challenges aplenty even when you're working
with strong primitives; trying to mix in weak stuff is just wasting
everyone's time.  People have encouraged me in the past to assume the
possibility of mere incompetence rather than assuming actual malice
(e.g. when the FBI's Louis Freeh testified to Congress about the security of
DES).  So: Were any of you on the standards committees for cellphone
privacy?  How about on the 802.11 "Wiretap Equivalent Privacy" committee?
Did any of you have a hand in shortening the key in DES?  Perhaps you
designed the encryption scheme used in DVDs or in Adobe eBooks?  Whether
you're incompetent or malicious, stick to breaking codes, it's much easier.
Especially when you break them in the standards committee before they're

Microsoft's PGP keys don't verify

<Brian McWilliams <>>
Thu, 26 Jul 2001 15:33:10 -0400

  [From Dave Farber's IP, archived at
  Submitted by Ben Laurie, who commented that
    As the immortal phrase has it, "the RISKS are obvious."

FYI ...

Microsoft Bulletins Fail PGP Verification

For at least four months, Microsoft has been sending out security bulletins
which fail a popular e-mail authentication system. As a result, the company
could be opening the door to counterfeit bulletins from malicious hackers.

To protect against forgery, Microsoft's security response center digitally
signs its bulletins with PGP before e-mailing them to subscribers of its
security notification service. But since at least March, if recipients
attempt to verify the messages' authenticity, PGP will issue a warning that
the bulletins contain an invalid signature.

"The problem is that Microsoft's bulletins effectively look as if they're
forged. And telling a Microsoft forgery from someone else's is virtually
impossible," said Paul Murphy, head of information technology at Gemini
Genomics, a genetic research firm in Cambridge, England.  [...]

Telling all to the police

<Norm <>>
Fri, 27 Jul 2001 18:06:01 -0400

*The New York Times* reports (27 Jul 2001) on 17 Jul theft from 9 lockers at
an upper East Side sports club.  Directly after they called the police a
call was received from "the police fraud department" and 4 victims responded
to a series of questions and gave their credit card numbers, husbands names,
SSN, PINs and mothers' maiden names.  Anything wrong with that?  That is,
aside from when the police did arrived they said there is no such dept.

One womans tale:  she called the credit card issuers but couldn't reach
her bank, being after hours and all.  The next morning she found $500
had been taken using her bank card.

The Risk, stupidity or cupidity aside, is being unlucky enough to be a
victim outside bankers hours ... and in a bank not having a 24-hour
notification phone#.  There Oughta Be A Law, as credit cards, that limits
consumer loss to $50 for such cases.

(PS: the same woman said she had worked out daily until then but "Now I am
so paranoid I haven't been back".  That's probably the wrong lesson

Norm deCarteret    NSDEC Inc

Identity theft

<Jack Holleran <>>
Fri, 27 Jul 2001 00:12:26 -0400

It would interesting to see what the vetting process was for the
salesperson(s)?  There seems to be an incredible amount of information that
was revealed without (m)any controls in place.

  Huge identity theft uncovered; Files with Social Security and driver's
  license numbers pasted in chat room; possible link to cell phone
  applications, By Bob Sullivan, MSNBC, 25 Jul 2001

  Key personal data belonging to hundreds of individuals have been shared in
  an Internet chat room, in what one expert says could become one of the
  largest identity-theft cases ever. The data include Social Security
  numbers, driver's license numbers, date of birth and credit card
  information - everything a criminal would need to open an online bank
  account, apply for a credit card, even create the paperwork necessary to
  smuggle illegal immigrants. It is still unclear how the data ended up in
  the chat room, but an investigation has revealed common threads
  among the victims - including the purchase of a cell phone online from or an AT&T Wireless reseller.
  Full text of the article can be found at

Risks of profanity filtering

<Paul Bissex <>>
Thu, 19 Jul 2001 20:36:19 -0400

Observant readers will have already noted that my last name contains the
word "sex." Recently, in trying to register with a Web site -- using my real
name -- I was chastised for "profanity" and asked to choose a different ID.

I declined, and since the company has offered no response to my inquiry as
to whether this policy is really necessary, I thought I'd share the screen

The business risk, alienating customers, is fairly obvious.  More broadly,
this highlights a familiar problem with "bad-word list" censorware. Imagine
if this were an e-mail filter on a firewall instead of a registration

Paul Bissex, CEO,

Car-door lock remote control activates another car's alarm

< (Mark Brader)>
Tue, 31 Jul 2001 17:51:01 -0400 (EDT)

  [The following was posted by "K.D." <> in; forwarded to Risks by Mark Brader with the
  author's permission.]

On at least three occasions, my battery-operated car unlock remote control
set off the car alarms in nearby vehicles.  I found that I could turn on the
alarm, and with another press of the remote control, turn it off.  Ad

The most astounding of these was the first time it happened (in Rochester,
NY).  But not simply because this was new information that I could set off
the car alarm.  Rather, because of the reaction of the owner of the other

At first, I didn't realize how it was that the nearby car alarm had been
triggered, since neither we or anyone else was close to the vehicle.
Eventually, I figured out that *I* had done it with my remote control.  I
also figured out that I could turn the alarm off as well as turn it on.

As we continued to approach my car, it also dawned on me that the owner of
the vehicle in question was also coming across the lot.  I pointed to the
previously alarm-sounding vehicle, and asked if it was his.  He said that it
was.  Lest he had observed what had just occurred and somehow thought I was
up to no good, I said, "It seems that my remote control activates your car
alarm."  His response?  "I don't have a car alarm."  I looked at my sister,
who was as perplexed as I was, and decided not to argue / explain.

Perhaps I should mention that there was no other vehicle nearby that could
have been sounding the alarm.  It was not my vehicle.  Especially now that
it has happened two other times, I am sure my remote control triggered his

One bummer about this is thus:  You open your car door.  The other person's
alarm goes off.  You press the remote control again to shut off the alarm
and naturally your car door locks again.  So, you have to unlock your car
door, the alarm of the other car goes off, you open your car door so you can
get in your car, and then you press the remote control again to shut off the
other car alarm.

I suppose if you just gave the alarm enough time, it would shut off on its

[K.D. then posted this followup]

In re-reading my post, I realized that this is screwed up.  Yes, when you
hit the remote control again ("unlock"), the door lock makes a sound.
However, I now realize that it isn't re-locking -- just making that
unlocking sound again.

I think.  At the time (second incident), I recall that when I tried to get
into my car after turning off the other person's car alarm, I found that my
door was locked.  Duh -- I had just unlocked it.  I then assumed that I had
relocked my car door, even though I had presumably used "unlock" to turn off
the alarm, as that is the button I had used that resulted in the alarm
turning on.

Damn -- if and when this happens again (so far, three times in about as many
months), I'll have to study the phenomenon more closely.


S-not-SL (Re: SSL, RISKS-21.53,54)

<Mike Albaugh <>>
Mon, 23 Jul 2001 16:07:25 -0700 (PDT)

I have found the following analogy useful, explaining to laypersons the
"Security policy" most common on the Web:

  "Imagine a restaurant that assigns armed guards to escort your credit-card
  to the cash-register and back, then tacks all the carbons to the
  employee-bulletin-board, right inside an un-locked back door"

Most of them get it immediately.

Re: MSN security upgrade forces new e-mail address (Silberman, R-21.54)

<"Robert J. Woodhead (AnimEigo)" <>>
Mon, 23 Jul 2001 19:52:39 -0400

"Ami A. Silberman" <> wrote:
>The risk? MSN's attempt to improve security (apparently by forcing spammers
>to modify their software to change fake msn addresses) has resulted in
>additional burden on list administrators.

You think that's bad?  I've been maintaining a bounce-management tool for a
mac-based listserv app, and as such, I see a lot of weird bounce formats,
many of which make the extraction of the bouncing e-mail address quite a

But the all-time champ is a certain large ISP who shall remain nameless but
whose initials are a,o & l.  If one of their users has redirected his e-mail
to another ISP, and the final destination e-mail address bounces, then our
friendly large ISP sends a polite bounce message back that clearly contains
the final destination e-mail address.

Alas, since it doesn't contain the original destination e-mail address, it is
impossible to determine who to unsubscribe; forevermore, you have a "zombie"
in your mailing list.

The listserv, alas, doesn't attach a useful header like
"Original-Recipient:" that could be used to identify the zombie because it
tries to conserve bandwidth by grouping e-mails to the same domain name into
a single transaction.

If mail servers added an "Original-Recipient:" header if they have to
forward the e-mail (and there isn't already one in the headers), life would
be immeasurably easier for bounce management. A standard for bounce
reporting that made life easy for nonhumans would also seem to be an obvious

Needless to say, an e-mail to the large ISP mentioning the issue seems
to have gotten sucked into a black hole.

Robert Woodhead, CEO, AnimEigo   The Net's only URL registration SHARESERVICE.

No Appleplexy needed (Re: RISKS-21.55)

<Dave Stringer-Calvert <>>
Tue, 31 Jul 2001 16:13:52 -0700

The Apple-DNS-hacked item in the latest risks is not a hack - it's a
"legitimate" use of the NIC records.  Someone has registered hosts with
the NIC who just happen to have in their name.  The same thing
has been done to Microsoft:

  ; whois

  [... and so on into the night]

     [This was noted by MANY readers.  TNX.  Sorry for my immoderate lapse.

Re: Autoresponder goes haywire (Bieber, RISKS-21.51)

Sun, 22 Jul 2001 11:21:43 -0600 (MDT)

In RISKS-21.51, Joshua M Bieber mentioned the following problems that led to
a quite typical autoresponder flood of one of his mailing lists.  In
addition to the suggested protective measures, it may also be wise for the
list to send with the original sender's address as the From/Reply address,
instead of using the list broadcast address there.  That way, only one
person gets nailed with the inappropriate autoresponse.  That's still
unacceptable behavior, but at least the damage is less severe that way.

Better yet, however, is deleting that bogus autoresponder software.  Any
autoresponder that replies to a message where:

    1) the Precedence header indicates Bulk or List, or where
    2) the user's address does -not- appear in the To or Cc header

is broken software.  The author of such an autoresponder should at least be
hauled out behind the barn for a strapping.

The problem illustrated by Guilty Person, in cahoots with the author of the
broken autoresponder used, is that of continually rewriting the same piece
of software with the same old mistakes.  In this case, it's particularly
ludicrous; properly operating examples of 'vacation' have been available
for free for 20 years.

Richard Johnson

Re: Erroneous air navigation reading (Hopkins, RISKS-21.53)

<Mike James <>>
Mon, 23 Jul 2001 21:54:24 +0100

That description of an aircraft navigation system out by 14 degrees on a
bearing posted by Bill Hopkins reminds me of a 'trick' I played on myself.

Take one Garmin GPS 12(XL),38,45,48 .... Probably most handheld GPS units.

Set user compass variation...

  setup->navigation->heading->user and enter 180 degrees

Now navigate to a waypoint. The bearing to waypoint will be displayed as
asked, but 180 degrees out. The 'compass' display arrow correctly
contradicts the bearing given.

This is confusing but totally correct. Just be careful....

Smaller numbers would be less obvious as in the aircraft case.

I was only yacht racing in the Solent and the error was obvious.  Crashing
off a mountain by using a magnetic compass and a GPS misconfigured like this
could be worse. (standing still need magnetic compass for current heading)

Re: Polarized sunglasses and LCD frustration (Boyd, RISKS-21.54)

<Chris J Dixon <>>
Fri, 27 Jul 2001 19:06:45 +0100

Surely it is the other way round.  Displays are not fussy about the
polarisation angle, but sunglasses are specifically oriented so that they
are most effective at intercepting light reflected off (and polarised by)
horizontal surfaces.

Chris J Dixon  Nottingham UK <>

Please report problems with the web pages to the maintainer