The RISKS Digest
Volume 20 Issue 29

Friday, 2nd April 1999

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…


o Attack of the Tuxissa Virus
o Computer crash creates nonpersons in Zurich
Bruce Walker
o tcpd warning
Kragen Sitaker
o Saving files on shared computers
Bertrand Meyer
o Self-opening car windows ...
Jeremy Folkes
o Swedish telephone outage
Danny Kohn
o Electricity over Internet
Lionel Cons
o In the summertime, when your VCR screws up
Michael Bacon
o Brain-dead PacBell automated payment promise system
Michael D. Crawford
o Re: Unusable backup power
Terry Harris
o Origins of PC / Mac Virus Vulnerability
Mich Kabay
o Re: More e-mail risks
Michael H Buselli
o Re: Apple Y2K
Art Delano
o REVIEW: "Information Warfare and Security", Dorothy Denning
Rob Slade
o Info on RISKS (comp.risks)

Attack of the Tuxissa Virus

Anonymous <nobody@REPLAY.COM>
Tue, 30 Mar 1999 05:31:34 +0200
  [Contributed belatedly by several RISKS readers.  TNX.  Sorry we could
  not have included it in one of YESTERDAY'S three issues.  PGN]

LART* Advisory LA-99.01.Tuxissa
Original issue date: Apr. 0a, 1999
Last revised: --

Topic: Attack of the Tuxissa Virus

This advisory is intended primarily for network administrators responsible
for user configuration and maintenance.

Attack of the Tuxissa Virus, March 29, 1999

What started out as a prank posting to comp.os.linux.advocacy yesterday has
turned into one of the most significant viruses in computing history.  The
creator of the virus, who goes by the moniker "Anonymous Longhair", modified
the well-known Melissa [1] virus to download and install Linux on infected

"It's a work of art," one Linux advocate told Humorix after he looked
through the Tuxissa virus source code.  "This virus goes well beyond the
feeble troublemaking of Melissa."  The advocate enumerated some of the tasks
the virus performs in the background while the user is blissfully playing

Once the virus is activated, it first works on propagating itself. It has a
built-in e-mail harvesting module that downloads all the pages referenced in
the user's Internet Explorer bookmarks and scans them for e-mail addresses.
Using Outlook, the virus sends a copy of itself to every e-mail address it
comes across.

After it has successfully reproduced, the virus begins the tricky process of
upgrading the system to Linux.  First, the virus modifies AUTOEXEC.BAT so
that the virus will be re-activated if the system crashes or is shut down
while the upgrade is in process. Second, the virus downloads a stripped-down
Slackware distribution, using a lengthy list of mirror sites to prevent the
virus from overloading any one server.

Then the virus configures a UMSDOS filesystem to install Linux on.  Since
this filesystem resides on a FAT partition, there is no need to re-partition
the hard drive, one of the few actions that the Word macro language doesn't

Next, the virus uncompresses the downloaded files into the new Linux
filesystem.  The virus then permanently deletes all copies of the Windows
Registry, virtually preventing the user from booting into Windows without a
re-install.  After modifying the boot sector, the virus terminates its own
life by rebooting the system. The computer boots into the Slackware setup
program, which automatically finishes the installation of Linux.  Finally,
the dazed user is presented with the Linux login prompt and the text,
"Welcome to Linux.  You'll never want to use Windows again.  Type 'root' to

The whole process take about two hours, assuming the user has a decent
Internet connection.  Since the virus runs invisibly in the background, the
user has no chance to stop it until it's too late.

The e-mail message that the virus is attached to has the subject "Important
Message About Windows Security".  The text of the body says, "I want to let
you know about some security problems I've uncovered in Windows 95/98/NT,
Office 95/97, and Outlook. It's critically important that you protect your
system against these attacks.  Visit these sites for more information..."
The rest of the message contains 42 links to sites about Linux and free

Slashdot is one of those links.  "That could spell trouble," one Slashdot
expert told Humorix.  "Slashdot could fall victim to the new 'Macro Virus
Effect' if this virus continues to propagate at its present exponential
growth rate.  Red Hat's portal site, another site present on the virus'
links list, seems to be quite sluggish right now..."

Details on how the virus started are a bit sketchy.  The "Anonymous
Longhair" who created it only posted it to Usenet as an early April Fool's
gag, a demonstration of how easy it would be to mount a "Linux revolution".
Some other Usenet reader is responsible for actually spreading the virus
into the wild.  One observer speculated, "I imagine the virus was first sent
to the addresses of several well-known spammers.  The virus probably latched
on to the spammer's e-mail lists and began propagating at a fantastic rate.
With no boundary to its growth, this thing could wind up infecting every
single Net-connected Wintel box in the world.  Wouldn't that be a shame!"

Linus Torvalds, who just left for a two week vacation, was unavailable for
comment at press time.  We have a strong feeling that his vacation will be
cut short very soon...


James S. Baughn

  [For those of you not familiar with the imagery, think about what erect
  short-legged flightless aquatic-bird operating-system symbol seems to be
  wearing a tux.  But then don't ask about who the Mel in Melissa is.  PGN]

Computer crash creates nonpersons in Zurich

"Walker, Bruce" <>
Thu, 1 Apr 1999 10:47:57 -0800
I found this one while surfing the web

One side effect of the year 2000 problem; crash in central computer

On Wednesday afternoon, 31 Mar 1999, the central computer of the Zurich city
administration crashed due to a mistake by two officials who were working on
the reprogramming the cantonal and city information systems for Y2K.  After
rebooting the computer, they found that important files were missing --
including those containing data on the citizens of Zurich, many of whom may
have become ``officially lost'', according to a city councillor.  [Source:
*Zurich Tages-Anzeiger*, 1 April 1999, PGN-ed from the German and from
Bruce's translation]

My comment is: where were the backup tapes?

Bruce Walker  1406 Stonewood Ct.  San Pedro, Calif. 90732

tcpd warning

Kragen Sitaker <>
Fri, 2 Apr 1999 17:32:51 -0500 (EST)
The problem with tcpd that's being discussed on BUGTRAQ — that a backslash
at the end of a comment line has a meaning different from the one a
particular user was expecting — illustrates a larger problem.

Simple, predictable user interfaces are essential to security.  If you must
direct a program to take some action, and you cannot tell whether the action
you have directed it to take is dangerous or not, then your system is
insecure, because there is a significant chance you will accidentally do
something dangerous.

You don't want DWIM in your hosts.allow and hosts.deny files.  You don't
want confusingly-difficult syntax, either.

In general, this is an argument against using computers to safeguard
security, because computers are hard to predict.  (Much harder, for example,
than deadbolts.)  But there are cases in which you have to do this; in these
cases, it is imperative to use software that is as transparent as possible,
so you can have a high degree of assurance that it will do the right thing.

qmail's config files are good in this way: very, very simple.  Very hard to
get surprising behavior — at least, surprising to me.  It is quite likely
that new Unix users would be surprised by it frequently, and I don't know
any way around that.

Usability is essential to security.

<>  Kragen Sitaker  <>

Saving files on shared computers

Tue, 30 Mar 1999 20:06:01 -0800
Mail from: Bertrand Meyer (Interactive Software Engineering, Santa Barbara)

Recently I used a public access computer graciously provided by HP in an
American Airlines Admiral's Club. Although a long-time RISKS reader, I
couldn't believe my eyes when looking under "Documents" in the Windows
taskbar. Previous users had stored all kinds of Word documents, including
one entitled "Confidential Settlement Terms", personal correspondence etc.

The service is extremely useful. I would be sorry if pointing out its risks
incited its providers to restrict it. Still, people who offer such shared
services should think of sticking on the computer a note of the form "When
you are done, remember not to leave behind any personal file".

-- Bertrand Meyer, ISE, Santa Barbara,

Self-opening car windows ...

Jeremy Folkes <>
Sun, 28 Mar 1999 0:20 +0000 (GMT Standard Time)
A few weeks ago I took delivery of a shiny new Ford Galaxy - a 7 seater
"people-mover".  About two weeks later I came out of the house in the
morning to find both front windows open - although I was convinced that I'd
shut it all up the night before.  My wife didn't really believe me, but then
it happened to her - she parked it up with the windows closed, and when she
returned to the car, the windows were opened.  She took it to the garage,
they didn't really believe her, until it happened to them....

In the good old days when windows were opened by turning a handle, they
tended to stay in the position you left them.  Now with electronics taking
over, I had a car which would randomly open its windows when left alone.
Luckily, nothing was stolen, but had I have been parked in a different
place, it could have been more serious - or at the least if it had been
raining I could have come back to a very wet car.

The Risks.  Technology may be convenient, but it brings all sorts of new
failure modes that we probably never envisaged.  The cleverer it gets, the
more opportunity for things to go wrong....


Swedish telephone outage

"Danny Kohn" <>
Thu, 25 Mar 1999 14:13:31 +0100
After a number of ISDN outages last year and some this year in the country,
our nationally owned telco Telia had two big outages in the capital of
Stockholm.  It happened the first time Monday 15 Mar 1999, when millions of
phone lines including the police headquarters' PBX were unusable for 8
hours!  The outage was repeated exactly a week later between 10:25am and
11:05am, when incoming calls to the police PBX and to another 250 business
PBXs where blocked.

The second outage is explained as an intermittent error that disturbed the
communication between PBXs and the telco equipment.  In addition the
software that would localize the problem had a bug so that the error would
not display.

The police management said that they trust Telia that this problem is now

Comming to mind is that telco exchanges are often purchased in international
competition.  A telco operator can not see through the software.  But given
the complexity neither can the producer — we might not have bugs if they
did.  So, if a intruder paid by some nearby country wanted to, he could
program some code "detonating" as a part of war attack.

Danny Kohn  Systematik Consulting AB +46(0)708 140300

Electricity over Internet

Lionel Cons <>
Tue, 30 Mar 1999 14:47:17 +0200 (METDST)
There's no doubt that the Internet relies on electricity, it seems that some
may have electricity relying on the Internet:

 In Latest E-Commerce Play, Venture Sells Electricity Online - Internet
World 3/22/99

This is a nice chicken and egg problem: which one will fail first?

Lionel Cons

In the summertime, when your VCR screws up

Michael Bacon <>
Wed, 31 Mar 1999 12:02:48 +0100
The UK changed to Summer Time (GMT+1 = BST) at 02:00 hours on Sunday 28
March 1999.

A national newspaper printed a warning in the TV listings, that VCR owners
should not adjust the clock on their VCR until 06:00 as alteration (to BST)
would invalidate the settings of the Video Plus+ system.  This is a system
whereby a code number is assigned to each programme (identifying channel,
start day and time, and duration).  This number an be entered into the VCR
(via the remote control) and it automatically records the programme(s) as

My VCR takes its clock from the broadcast Teletext system and adjusts the
time automatically.  As warned, any programming of the Video Plus+ system
would thereby be invalidated.  The impact of this RISK is minor, but extend
the application ...

Brain-dead PacBell automated payment promise system

"Michael D. Crawford" <>
23 Mar 1999 23:19:31 GMT
I just got a recorded message telling me to call Pacific Bell at an 800
number about my overdue bill.

I've had rather amazing phone bills for the last year because my girlfriend
lives in another country.

I sent them $1100 on March 20, so I figured I could call and just say the
check was in the mail.  Today's the 23rd.

After navigating through the maze, entering my phone number and the last four
digits of my social security number, I was told that I could send the bill in
two parts.  I selected the option to mail it.  Then when asked to enter the
date I would be mailing it as 4 digits, I entered 0320.  The machine asked if
I meant meant March 3, 2000?  No, I meant three days ago.  Entering 0324 was
acceptable to the machine, which is apparently unable to conceive of the
concept "the check is in the mail".

Then it said I needed to mail in $164.88.  I entered that I would send in
$200, just because I couldn't remember how much I needed to send in.  It said
that was unacceptable, the minimum due was $164.88.  That time I remembered
it and entered in $164.88.  It said that was unacceptable, the minimum due
was $164.88, so I hung up.

The amount that's actually past due is $900, so that will be covered when the
check is received.  I figure if the money's important enough for them,
they'll call me with a real human about the arrangements.
Michael D. Crawford
GoingWare - Expert Software Development and Consulting

Re: Unusable backup power (Weingart, RISKS-20.24)

"Terry Harris" <>
Fri, 2 Apr 1999 11:30:59 -0600
This is a multi-part message in MIME format.

Some years ago a friend worked at a site with an IBM 370/145.  The site
handled DP for several hospitals.  They decided they wanted a backup power
source to prevent interruptions in service.  They had a battery based system
installed that would give them several hours of operation.  The backup
system produced 208V 3 phase current at 60Hz.

It turned out that they and the backup system supplier had forgotten one
thing.  The 370/145 actually ran on 208V 3 phase 400Hz.  The 370 included a
motor-generator set to convert the current (a 60Hz motor connected to a
400Hz generator).  These motor generator sets had a considerable startup
current (initially about 375 Amps decreasing to about 20 over a few seconds)
and the backup system could not supply that much.

If the backup system switched in fast enough that the generator did not spin
down the computer was still useful.  If the backup power switch wasn't fast
enough it did no good as the computer could not be restarted.

The moral: understand all the requirements that the backup system will have
to meet.

Origins of PC / Mac Virus Vulnerability

Mich Kabay <>
Wed, 31 Mar 1999 16:58:47 -0500
Letter to the Editors, ATLANTIC MONTHLY.

Dear Editors:

I enjoyed Robert Buderi's historical overview of the rise of computer
viruses (April _Atlantic_) — a particularly timely contribution to public
knowledge given the explosive spread of the Melissa virus and similar
e-mail-enabled nuisances in late March 1999.

Unfortunately, technologically unsophisticated readers may have concluded
that computer viruses are inevitable outgrowths of the nature of computers.
Now, any author must select what to include or not, especially when writing
within a word-count limitation.  As a contribution to the discussion, not as
a criticism, I did want to add a couple of fundamental points that were not
mentioned in the article but that may help readers understand why we are
suffering this cyberplague.

1) Computer viruses in the wild exist almost exclusively on Microsoft DOS,
Microsoft Windows 3.x, 95, & 98 and Apple MacOS operating systems. For all
practical purposes, there are no viruses in UNIX, MVS, VMS, MPE and other
operating systems that run on workstations, minicomputers and mainframe
computers.  The root (no pun intended) of our PC virus woes is design
decisions made long ago.  Microsoft engineers decided to dispense with a
security kernel — the part of the operating system that determines exactly
what kinds of operations an ordinary user is permitted to perform.  The
security kernel limits certain other operations to supervisory users (often
known as "root") and yet others to the operating system itself.  In
well-behaved operating systems, changing a memory location (for example) is
usually a restricted operation.  Under Windows95, in contrast, every user is
in effect root and programs the user runs can do anything at all.  Without
this vulnerability, viruses in the PC and Mac worlds would be severely
limited in their effects (the "payload").

When PC and Mac operating systems introduce a security kernel into their
operating systems, much of the viral payload would be defused.

2) Microsoft engineers decided to turn office software into a general
programming tool.  MS-Office programs include macros — not an unusual
feature for word processors and spreadsheets in the last 20 years.
Unfortunately, they also decided to allow two functions that made their
software vulnerable to macro viruses: (a) automatic execution at startup;
and (b) access to system routines.  Without these features, Word and Excel
macro viruses would not be as powerful nor as virulent in their replication
as they are today.

When Microsoft changes the defaults on Word and Excel to prevent automatic
execution of macros, PC and Mac viruses will have greater difficulty in

Public awareness of the design decisions underlying our vulnerability may
help sway software engineers towards improvements in our working

M. E. Kabay, PhD, CISSP, Director of Education, ICSA Inc. — Vermont
255 Flood Road  Barre, VT 05641-4060 <>  802-479-7937

Re: More e-mail risks (Brown, RISKS-20.28)

Michael H Buselli <>
Fri, 02 Apr 1999 11:21:31 -0600
> 1.  If too many people use this thing, we may get spamming incidents in
> which the spammer adds a PostPet attachment to the spam, ...

I think risk readers should know that Microsoft Outlook does this kind of
thing now.  If you want to get a guaranteed reply when someone using Outlook
as their mail client looks at your e-mail, add the following header:

    Return-Receipt-To: Michael H Buselli <>

Now, I've noticed that Outlook ignores the address given by this header
and uses the address specified by the "Sender:" header instead, but
that's not difficult to set to be whatever you want.

I used Outlook 98 to perform my tests playing with this header.  I suspect
that all other versions do the same.  I am not familiar enough with Outlook
Express to know if the risk applies to that mail client as well.

Perhaps there should be an option in Outlook to confirm or turn off these

Michael H Buselli ||

Re: Apple Y2K (Stringer-Calvert, RISKS-20.28)

Art Delano <>
Fri, 2 Apr 1999 09:33:33 -0500
>                 "We may not get everything right,
>        but at least we knew the century was going to end."
>    — Apple Computer, HAL 9000 ad for Macintosh Y2K compliance

To the best of my knowledge, that wasn't used in the Apple's Superbowl
commercial (the commercial is available online for download, but i feel
lazy). On the other hand, the same quote had been circulating, attributed to
Douglas Adams, for at least half a year.

A quick check with a search engine didn't turn up the original context of
the quote, but did find Apple's citing Adams on their own website: — as an introduction to their own Y2K
compliance statement.

Incidentally, Apple's statement that currently-supported hardware and
software are Y2K compatible allows them to make preemptive decisions to
discontinue support for any products they may decide are better dropped than
revised. While they're hardly the only company to do this, they might at
least provide documentation of what has been discontinued and recommend
upgrade paths, to keep their customers from guessing at what they need.

As has already been observed, even if a supported system, as shipped, will
be Y2K-compatible, third-party software may not be. Users may not be willing
to spend money and time and effort replacing working software with other
working software that provides no apparent outward improvements. I have
doubts that the average computer-using consumer has a clear understanding of
the distinctions between disparate applications, utilities, and system
functions, particularly in our current era of 'seamlessly' integrated
software, and will blame all problems at whichever company seems to have
their name most prominently displayed on the package.  The confusion is
probably better left as a subject for another thread.

  [Also noted by  DeRobertis <>.  PGN]

REVIEW: "Information Warfare and Security", Dorothy Denning

Rob Slade <>
Tue, 23 Mar 1999 08:44:02 -0800

"Information Warfare and Security", Dorothy Denning, 1999,
0-201-43303-6, U$34.95/C$52.50
%A   Dorothy Denning
%C   P.O. Box 520, 26 Prince Andrew Place, Don Mills, Ontario M3C 2T8
%D   1999
%G   0-201-43303-6
%I   Addison-Wesley Publishing Co.
%O   U$34.95/C$52.50 800-822-6339 Fax 617-944-7273
%P   522 p.
%T   "Information Warfare and Security"

Denning has chosen to take an inclusive approach to the topic of
information warfare, not limiting the material to attacks on
"military" targets.  Given the state of physical warfare, this seems
to be quite realistic.  It does mean that the book tends to read like
a high level computer security text (small wonder) with an emphasis on
intrusions and the more overt aspects of computer crime.

Part one is a foundation and background for the material to come.
Chapter one looks at the great many information aspects to the Gulf
War and Operation Desert Storm.  One of the unusual factors reviewed
is that of propaganda, or "perception management."  A theory of
infowar is the intent of chapter two, which outlines players and
positions in a variety of ways.  The theory is somewhat weakened for
being strongly dependent upon the idea of the value of the information
being attacked or defended, and this is an area that still requires
work.  Another possibly problematic area is the reliance on a "win-
lose" model for data warfare, when there have been numerous instances
of intruders, upon sufficient provocation, being willing to deny
themselves a resource by damaging it, on the basis that the defenders
stand to lose far more.  (On the other hand, "bragging rights" seem to
have a lot of value in the computer underground.)  More detail on the
players involved, and the possible types of attacks that have
occurred, and might occur, are presented in chapter three.

Part two looks at the specifics of offensive information warfare.
Chapter four is extremely interesting, showing that "open source," or
publicly available information, can and has been used for offensive
and criminal undertakings in a variety of ways.  Disinformation is
reviewed in chapter five, including the odd phenomenon of urban
legends and Internet hoaxes.  The problem of damage from insiders,
including, finally, a documented case of a salami attack (albeit a
rather clumsy one), is covered in chapter six.  Chapter seven
discusses the interception of information and communications in a
variety of ways, and, as a sideline, jamming and alteration.  A
variety of methods of computer intrusion are presented in chapter
eight.  False identity, both  identity theft and outright false, are
examined in chapter nine.  The material on viruses and worms, in
chapter ten, is solid, although I was sorry to see that a great many
possibilities for reproductive mayhem that have been discussed over
the years went unmentioned.  ("Harlie," Dr. Denning.  "When *HARLIE*
Was One.")  (Of course, when I sent the first draft, I had, myself,
spelled "Harlie" incorrectly.)

Part three looks at the opposite side, that of defence.  Chapter
eleven gives a good background to encryption, but, seemingly,
primarily as a general concept, rather than going into detail on
specific uses for protection.  Authentication is dealt with in chapter
twelve, and uses some of the cryptologic background.  With monitoring
and detection bracketing chapter thirteen, the section on firewalls
seems just slightly misplaced.  Chapter fourteen looks at risk
analysis, planning, and some resources.  The final chapter discusses
defence of the nation, and national policy in this regard, with
particular emphasis on the current situation in the US.

The content of this book not only presents a clear picture of a number
of aspects of information warfare, but does so in a very practical
manner, informed by the need to use "real world" examples.  In
addition, the anecdotal evidence backing the material makes the book
quite readable and interesting.  As a text for a course in information
warfare, it is complete and solidly based.  As a reference for
security analysts and practitioners, it is clear and thought-
provoking.  For those who may merely have some interest in the topic,
it is engaging and informative.

copyright Robert M. Slade, 1999   BKINWRSC.RVW   990212    or

Please report problems with the web pages to the maintainer