The RISKS Digest
Volume 25 Issue 34

Sunday, 21st September 2008

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…


SciAm article on Smart Grid
William P.N. Smith
Wall Street; where nothing can go worng wrogn wrgno....
David Lesher
Mortgage loan crisis due to wishful thinking, Garbage In Garbage Out
Geo Swan
BNY Mellon data breach now at 200K in Mass, 12M in U.S.
Monty Solomon
Risks of financial systems too complex to understand
Daniel P. B. Smith
Risks of not using check digits in bank account numbers
Toby Douglass
Risks of banking in Holland
Toby Douglass
Re: PayPal phishes their own customers
Sidney Markowitz
Re: Automated Bill Payments Are a Cinch: Not So Fast
Capability creep strikes again
Jay R. Ashworth
Expiration of cryptographic certificate killed airline ticket
Kenji Rikitake
Antivirus software in critical systems?
Martyn Thomas
Re: Antivirus software in critical systems? Aurora!
Al Mac Wheel
Re: Control-Z vs. Bourne-Again SHell
Re: Risks of GPS Devices ...
Richard Grady
USENIX Annual Tech '09 Call For Papers
Lionel Garth Jones
Info on RISKS (comp.risks)

SciAm article on Smart Grid

<"William P.N. Smith" <>>
Fri, 19 Sep 2008 09:17:15 -0400

I was just reading a great article on the need for a Smart electric Grid at but
found myself hoping that the people designing it are Risks readers.

Running the grid on a fly-by-wire system, and getting better efficiencies
out of it by running it closer to the edge of stability are pretty scary
concepts for long-time Risks readers.  The thought of replacing every device
(switch, circuit breaker, etc.) with a computerized version makes my brain
want to explode.

Not that it _can't_ be done, any more than the ATC system can't be upgraded,
but exhortations to do so make me want to ensure I'm current on my generator

Wall Street; where nothing can go worng wrogn wrgno....

<"David Lesher" <>>
Thu, 18 Sep 2008 18:06:10 -0400 (EDT)

Saul Hansell's blog, How Wall Street Lied to Its Computers, 18 Sep 2008,

I called some old timers in the risk-management world to see what went
wrong.  I fully expected them to tell me that the problem was that the
alarms were blaring and red lights were flashing on the risk machines and
greedy Wall Street bosses ignored the warnings to keep the profits
flowing. ....

But she and others say there is more to it: The people who ran the financial
firms chose to program their risk-management systems with overly optimistic
assumptions and to feed them oversimplified data. This kept them from
sounding the alarm early enough.

a) In other words, "Tell ourselves what we want to hear..."
b) Quis custodiet ipsos custodes

Mortgage loan crisis due to wishful thinking, Garbage In Garbage Out

<Geo Swan <>>
Fri, 19 Sep 2008 10:14:04 -0700

Saul Hansell, writing his *The New York Times* blog on 18 Sep 2008, offered
a detailed account of how the mortgage loans crisis was due to wishful
thinking in the entry of data to the automated tools bankers now count on to
warn them of risky loans.

While the principle of "Garbage In, Garbage Out" is old news to RISKS
readers, I thought this article was a useful reminder that the risks of GIGO
still has not penetrated to the general public.

BNY Mellon data breach now at 200K in Mass, 12M in U.S. (Re: R-25.31)

<Monty Solomon <>>
Sat, 20 Sep 2008 09:55:18 -0400

Continuing the saga re: BNY Mellon noted in RISKS-25.31, earlier reports
indicated that perhaps 200,000 Massachusetts consumers and 4.5M people
nationwide may have been affected.  The corrected numbers now exceed
400,000, in Massachusetts and 12,000,000 nationwide according to the NY
state Attorney General Martha Coakley.  [Source: Jonnelle Marte, Bank's data
loss may hit 400,000 Mass. customers, *The Boston Globe* online, 19 Sep
2008; PGN-ed]

Risks of financial systems too complex to understand

<"Daniel P. B. Smith" <>>
Wed, 17 Sep 2008 10:16:09 -0400

*The Boston Globe*, 17 Sep 2008:

  "_Bottom to crisis nowhere in sight_. The root of the panic on Wall Street
  is that investors have lost $350 billion from their securities backed by
  subprime mortgages. But on top of that, Wall Street holds tens of
  trillions in other arcane securities that were based on these
  mortgages. These mortgage securities are so complex that it's hard to
  calculate the value of these investments because owners do not clearly
  disclose their holdings. As a result, analysts and economists can't grasp
  how widespread the problem is... [A professor said] 'You know there are
  securities linked to the value of mortgages. You don't know how much those
  securities are worth...' Any attempts to gauge the size of these markets
  are 'an estimated guess,' said [an analyst]. 'When you zero in on the
  number, we don't know what the real risk is in there. It could be small or
  large. We just don't know," he said. "It's breathtaking.' [An economist
  said] 'it's hard to figure out who owes what to whom, and when that
  becomes a problem, markets don't function properly.'"

  "Wall Street firms... packaged mortgages together in pools, issued
  securities backed by the pools, and then sold them to investors. In
  addition, they created and traded complex securities known as 'credit
  default swaps...' The insurance contracts form an unseen but tangled web
  of connections linking AIG and Wall Street investment banks to hedge funds
  and even countries that invested in them. They are also one item in a
  smorgasbord of similar contracts - many of them based on corporate debt,
  credit cards, automobile loans, and commercial property loans - that is an
  unfathomable $63 trillion."

Plenty of blame to go around, but where do computers come in?

Data processing made it possible to manage mutual funds with more than a few
dozen stocks in them, like the first S&P index fund.  Computer networking
made it possible to interconnect financial products from different companies
and create multilevel financial products, like Fidelity's Dynamic Strategies
Fund. This "fund of funds" invests, not in stocks, but in other mutual funds
and ETFs...  _over a hundred of them_.

What would happen if you asked Fidelity for an inventory of the certificate
numbers of all of the stocks in that fund?

Computer technology made it possible to build multilevel pyramids of
financial products. These products contain many so investments, all many
steps removed from the product's owner, that the owner cannot possibly have
an intuitive grasp of them, or make any kind of reality check on their
risk. They are so distant from the actual investments that underlie them
that last November a federal judge in Ohio dismissed 14 foreclosure cases
brought on behalf of mortgage investors, because the investors were unable
to prove they actually owned the properties.

I once considered driving out to take a look at one of the commercial
properties whose address was listed in the TIAA Real Estate Fund's quarterly
report, just to see whether it was really there. I doubt that such a thing
has even crossed the mind of the people who trade credit default swaps.

It might turn out that, without anyone really noticing, the global financial
system has gradually turned into something that is mostly a MMORPG.

Risks of not using check digits in bank account numbers

<"Toby Douglass" <>>
Tue, 16 Sep 2008 11:10:40 +0200 (CEST)

Re: Automated Bill Payments Are a Cinch: Not So Fast, Erling Kristiansen

It should be statistically practically impossible to accidentally enter a
valid account number.  If account numbers have no check digits, the bank has
made a fundamental design error.

I have been living in Amsterdam now for four months (previously I lived in
the UK).  In the UK, my banks accounts are described by an eight digit
account number and a six digit sort code (I believe this is true for all UK
banks).  There are no check digits, but the account number must match the
sort code, which provides a useful validation.  In Holland, an account with
ABN AMRO is described by a single nine digit number.  There are no check
digits.  If you enter the wrong number, the bank will accept it, since it
cannot detect your mistake.

Risks of banking in Holland

<"Toby Douglass" <>>
Tue, 16 Sep 2008 12:03:47 +0200 (CEST)

Four months ago, I emigrated from Cambridge, England to Amsterdam, Holland.

Emigration requires at the destination the resumption of the organised
necessities of life - such as, for example, a bank account.

There are three main banks in Holland - ABN AMRO, Postbank and Rabobank.
(My commiserations to any banking entity not making my list).

On the basis of an almost nonexistent set of reviews for these banks and
their services, I chose ABN AMRO.

Fast-forward nearly four months to this most recent weekend.

I unfortunately suffer from a hopefully transient neurological deficit which
means my working memory is seriously impaired.  (Something that happens
naturally, it seems to me, to people as they move toward the slower realms
of advancing age, so I think I am not alone in this deficit.)

As such, I forgot my Dutch PIN.

I wanted to phone ABN AMRO, to ask them to change my PIN.  Unfortunately,
where the rental market is Amsterdam is nonexistent (State intervention in
the market has eliminated the market), I have not yet settled into a place
of my own and so do not have a fixed land line - only a Dutch SIM mobile.

ABN AMRO offer two customer service numbers; one for use inside Holland, one
for use internationally.  The internal number cannot be called from my
mobile.  I do not know why.  The international number detects you are
calling from Holland and after a short, polite, Eddiesque message - "you're
calling from Holland!  I have to disconnect you immediately!  share and
enjoy!" *click* - and so I cannot phone the bank.

This is of course a risk; don't needlessly eliminate redundancy in your

So, I have to travel, physically, to a branch.  Fortunately, I'm young and
healthy in body and traveling is not an issue.

I arrive at the bank — on Saturday, naturally, since I do not wish to take
a day off work just to visit.

I queue for a while.  Dutch banks like that.  I'm eventually approached by a
Rutger Hauer look-a-like.  I ask to change my PIN.

"Oh no,'re thinking you can change your PIN?  no."

"I can't change my PIN?  but then how do I make my card work again?"

"All we can do is issue you a new card.  It takes three days.  There's a

I note in the UK, card issuing is free and you can change your PIN over the
phone.  Perhaps if card issuing here was free, you would also miraculously
suddenly be able to change your PIN.

"Well, I need some money now - I don't have any cash on me and I need to do
shopping and I can't use my card."

"Oh no,'re thinking you can withdraw money?  no."

"I can't withdraw MONEY?!?"

"Not on the weekend.  You have to come during the week.  To this branch or
one of two others in Amsterdam.  No other branches provide money.  And no
branches provide money on the weekend."

(At this point, Rutger was looking at me as if I was slightly insane.
Clearly, I had picked up in the UK some dangerous Anglo-Saxon notions of
banking service, along the lines of being able to withdraw my money from my

"Look, I need to pay my rent.  Can I transfer some money?"

"Oh no,'re thinking you can transfer money?  no."

"I can't transfer money!?"

"No.  You can only transfer money on-line and you must have your card and
PIN.  Also, there's a charge."

"So...I can't withdraw money and I can't transfer money?"

"That's right."

(I looked around, warily, to check I had not fallen into a re-enactment of
The Castle and was being mistaken for the protagonist.  Alas, it was just an
ABN AMRO branch.)

"Okay, let me think...okay - if I closed my account, it has money in it now,
what would happen to that money?  you'd have to send it somewhere?"

"Yes.  If you close your account, we will transfer the money to the account
you specify.  It'll take a few weeks."

"A FEW WEEKS??!?!?"

"Yes.  The account is closed immediately of course, but there are checks and
so on that we need to do.  There is much bureaucracy that must be done. It
takes a few weeks before the money is transfered."

At this point I realised that I was reflexively clenching and unclenching my
fist and it seemed a prudent time to depart, lest the headlines --
"Embittered AMRO customer slays ten in money withdrawal horror!"  "AMRO says
-- we never expected people to forget their PIN!"

RISKS?  single point of failure upon an entirely plausible failure mode
(losing the PIN) with no fallback behaviour.  If for any reason you forget
your PIN, you can find yourself absolutely shut out of your account --
unable to withdraw money, unable to transfer money and even closing the
account, the final recourse, is arranged such that your funds are
sequestered.  You literally have no access to your money.  Like most people,
I need to eat and get grumpy when I'm starved for two days.

Furthermore, I now have the choice of paying a charge for a new card, which
I will use exactly once, to transfer my funds to my new Rabobank account
before closing my ABN AMRO account, or actually physically withdrawing a
couple of thousands euros and *walking* it across to my new bank to deposit

BTW, depositing cash into a Dutch bank?  there's a charge.

Re: PayPal phishes their own customers (RISKS-25.33)

<Sidney Markowitz <>>
Tue, 16 Sep 2008 10:24:23 +1200

A web search for SECURE.UNINITIALIZED.REAL.ERROR.COM shows that this is
neither phish nor phowl, it's a bug, not a pheature, though perhaps a bug
most phoul. Apparently an error message about an uninitialized variable is
being substituted where a URL is supposed to go in a template for the

This has been happening at least for almost a year.

Luckily whoever owns has not set up a phishing website to take
advantage of the bug. But the bug may not be as benign as simply producing a
bogus URL in the e-mail. This report indicate that it may be associated with
PayPal transactions that the user aborts being completed anyway:

This one is even worse. Not only does it seem to be associated with a bogus
charge to the PayPal account, but it demonstrates the bug showing up as bad
links in e-mail from a presumably live PayPal customer service representative
sent to a customer:

Sidney Markowitz

Re: Automated Bill Payments Are a Cinch: Not So Fast (RISKS-25.32)

<Huge <>>
Thu, 18 Sep 2008 10:22:56 +0100

I'm surprised no-one has mentioned the Direct Debit scheme which operates in
the UK, mostly very successfully;

  "Over 75% of adults in the UK have at least one Direct Debit and around
  100,000 organisations use Direct Debit for collecting a variety of regular
  and occasional bills including utility payments, insurance, council tax,
  mortgages, loans and subscriptions."

The "Direct Debit Guarantee" covers your rights;

"If an error is made by the organisation or your bank or building society,
you are guaranteed a full and immediate refund from your branch of the
amount paid."

To our (UK) eyes, US retail banking is very old-fashioned (*). I haven't
written a cheque (US: check) for a utility bill (or any other regular
payment) for years. Indeed, cheques are fading from use at an accelerating
rate and a number of large organisations no longer accept them. Cheque
processing was out-sourced by most UK retail banks some years ago as no
longer economic to keep in-house.

(* Given recent events, I am no longer sure if this is a good or bad thing.)

Capability creep strikes again

<"Jay R. Ashworth" <>>
Wed, 17 Sep 2008 16:35:18 -0400 (EDT)

Of course, it's being sold as a plan to Protect The Chilluns:

  Private companies in the US are hoping to use red light cameras and speed
  cameras as the basis for a nationwide surveillance network similar to one
  that will be active next year in the UK. Redflex and American Traffic
  Solutions (ATS), the top two photo enforcement providers in the US, are
  quietly shopping new motorist tracking options to prospective state and
  local government clients. Redflex explained the company's latest
  developments in an August 7 meeting with Homestead, Florida officials.

  "We are moving into areas such as homeland security on a national level
  and on a local level," Redflex regional director Cherif Elsadek
  said. "Optical character recognition is our next roll out which will be
  coming out in a few months — probably about five months or so."

  The technology would be integrated with the Australian company's existing
  red light camera and speed camera systems. It allows officials to keep
  full video records of passing motorists and their passengers, limited only
  by available hard drive space and the types of cameras installed. To gain
  public acceptance, the surveillance program is being initially sold as an
  aid for police looking to solve Amber Alert cases and locate stolen cars.

I don't, at the moment, see any way around this problem: corporations are in
it only for the money; if they can make that money by selling to governments
capabilities that it would be illegal for the governments to implement on
their own, they will.

It's a slippery slope, in the reverse of the usual direction.

And remember: if that database exists, your wife's divorce attorney will be
able to subpoena it.

Nice to know Al Queda doesn't need to take away our freedoms; like the
citizens of the Weimar Republic, we're doing a bang up job of it ourselves.

Jay R. Ashworth, Ashworth & Associates, St Petersburg FL USA  1 727 647 1274

Expiration of cryptographic certificate killed airline ticket

<Kenji Rikitake <>>
Thu, 18 Sep 2008 21:47:40 +0900

All Nippon Airways (ANA) made a public announcement on September 19,
2008 that the check-in system trouble for the domestic flights happened
on September 14, 2008 was caused by expiration of the cryptographic
certificate issued for access authentication of check-in terminals for
the ticket agents.

The company told that:
* the certificate was expired on September 14, 2008;
* the expiration was direct cause of the errors of the check-in
  terminals when the terminals were booted on September 14, 2008;
* the company knew when they first installed the certificate on 2005;
* they didn't activate the certificate until September 2007; and
* no one in the company did notice the expiration during the development
  of the terminal subsystem.

Sources: ANA public announcement (in Japanese)

The risks are obvious:

* If you depend on something which will expire, such as PKI certificates and
  domain name subscription, failure of the renewal may kill the whole
  subsystems depending on the expired objects;

* If you depend on something which will expire, a warning system (of
  pre-expiration notice) should be installed to tell you about the
  expiration, or you will forget about it and may cause the system trouble;

* You should always be aware of the *unused* functions of a system and how
  they work, especially when you decide to activate some of them.

Antivirus software in critical systems? (Re: PGN, RISKS-25.33)

<Martyn Thomas <>>
Tue, 16 Sep 2008 09:32:28 +0100

Self-evidently, fitness for purpose should trump cost of development in this
discussion, because the incremental cost of carrying out the final stage of
development (and then operation) in an environment that does not require the
system to include its own protection against virus infection is fairly
small, and the examples that have been given describe systems that are
clearly not fit for purpose as implemented.

On the other hand, if cost is more important to you than fitness for
purpose, I have a system waiting for you that costs 10% less than the lowest
estimate anyone else has given you. Give me a call.

Martyn Thomas CBE FREng

Re: Antivirus software in critical systems? Aurora!

<Al Mac Wheel <>>
Tue, 16 Sep 2008 12:19:15 -0500

Aurora is US Gov name for the mischief hackers' malware can do when targeted
directly at electric utilities.  Here's a link to a couple of recent Aurora

(a) $ 1 million generator blown up by hacking = DHS proof of concept.
(b) Vancouver Canada insider crime as union negotiating

US Congress hearings last week into what it takes to protect against this.
(a) Public Utilities complain that by US making Aurora details top
    secret, even from their industry, they cannot mitigate against all
    threats known to gov.
(b) US gov agencies say solution is for Congress to give them unfettered
    authority to meddle in industry, with zero accountability for how that
    authority is used.

Here's a link to those hearings.

Re: Control-Z vs. Bourne-Again SHell (jidanni/chau, RISKS-25.31-32)

Wed, 17 Sep 2008 07:17:59 +0800

The problem with commands like
$ sleep 666; destroy_planet
is that Dr. Evil can type them on your VT100 terminal and walk
gleefully away, while you try to find the bash page in the manual
racks to see if there are any safe control or break keys you can press
short of yanking the terminal out of the wall as time runs out... but
it was all a bad 1980's dream, as the librarian tugs your shoulder
"Sir, your half-hour is up and there are a lot of people waiting to
use a terminal."

Re: Risks of GPS Devices ... (Brader, RISKS-25.32)

<Richard Grady <>>
Tue, 16 Sep 2008 12:02:12 -0700

This is standard behavior on most amateur radio transceivers, because there
are many more functions needed than there are keys on the device.

One of my transceivers (Yaesu FT-8100R, a mobile 2-meter/70-cm unit) triples up
on some keys:
1. Press the key to activate the Normal function.
2. Press the F/M key (similar in concept to a shift or control key) and then
   press the key to activate the Alternate function.
3. Press the F/M key for 1/2 second or more and then press the key to activate
   the Super-Alternate function.

I have memorized some of the Alternate functions, but I usually have to
refer to the manual for the Super-Alternate functions.

  [Incidentally, in RISKS-25.34, Sergei apparently completely missed the
  humor that Mark Brader had very amusingly noted ("to be held down"), which
  my comment highlighted to make sure our readers did not miss it.  However,
  Sergei apparently did miss it.  I seem to have compounded the silliness by
  evidently oversimplifying the repetition of the gaffe that Mark quoted by
  trying to correct it, because Sergei had responded not to the humor that
  had amused Mark and that had inspired me to run Mark's item in the first
  place.  If that perhaps mistakenly gave anyone the impression that Mark
  would have disagreed with Sergei, I profoundly apologize to Mark.  This is
  a case where the humor should have been obvious to readers of Mark's item
  in RISKS-25.32, not the details of the implied corrected statement.  PGN]

USENIX Annual Tech '09 Call For Papers

<Lionel Garth Jones <>>
Wed, 17 Sep 2008 11:37:46 -0700

2009 USENIX Annual Technical Conference
June 14-19, 2009, San Diego, CA
Paper Submissions Deadline: January 9, 2009, 11:59 p.m. PST

On behalf of the 2009 USENIX Annual Technical Conference program committee,
we request your ideas, proposals, and papers for tutorials, refereed papers,
and posters.

Authors are invited to submit original and innovative papers to the Refereed
Papers Track of the 2009 USENIX Annual Technical Conference.  Papers can be
either full papers of at most 14 pages or short papers of at most 6
pages. Authors are required to submit papers by 11:59 p.m. PST, Friday,
January 9, 2009. (Note new deadline.)

In full papers, we seek high-quality submissions that further the knowledge
and understanding of modern computing systems, with an emphasis on
implementations and experimental results. Short papers should describe early
ideas, advocate a controversial position, or present interesting results
that do not require a full-length paper. We encourage papers that break new
ground or present insightful results based on practical experience.

The USENIX conference has a broad scope. Specific topics of interest include
but are not limited to:

--Architectural interaction
--Cloud computing
--Deployment experience
--Distributed and parallel systems
--Embedded systems
--Energy/power management
--File and storage systems
--Mobile, wireless, and sensor systems
--Networking and network services
--Operating systems
--Reliability, availability, and scalability
--Security, privacy, and trust
--System and network management and troubleshooting
--Usage studies and workload characterization
--Web technology

More information on these and other submission guidelines is available
on our Web site:

On behalf of the 2009 USENIX Annual Technical Conference organizers,
  Geoffrey M. Voelker, University of California, San Diego
  Alec Wolman, Microsoft Research

2009 USENIX Annual Technical Conference Program Co-Chairs

Please report problems with the web pages to the maintainer