WARNING—LONG! NTSB on WMATA Crash [DCA-09-MR-007] (Risks-25.79) On 27 July, 2010 NTSB held a public hearing to announce their report on the 22 May 2009 Metro crash of train #112 into #214. The report is technical, long, detailed, and leaves few parties untainted. Foremost, NTSB found the primary cause was "a failure of the track circuit modules, built by GRS/Alstom Signaling Inc., that caused the automatic train control system to lose detection of train 214"... The second cause was WMATA's failure to use tools they created after another detection failure/near miss in 2005. There are many more aspects detailed. I urge anyone involved in RISKS-topics to invest the time in reading the full report [final version due out soon] and exhibits, but I'll attempt a synopsis. Even it won't be short. WMATA uses bog-standard AC track signaling, around for over one hundred years. Basically, an audio frequency signal is injected across the rails [by a transmitter module] at one end of a several hundred foot block of track; if it's detected [receiver module] at the far end [receiver module], the block is empty & an output relay is held energized; if it is NOT, either a train is shorting Rail 1 to Rail 2, or something has failed; that track is considered occupied. In short, it should be fail-safe; broken wires, cracked rails, blown fuses etc all drop the relay. From those core inputs, the automatic train control system manages trains to maintain safe & desired spacing. Even in "manual" mode; the train operator can not do better than slowly creep into an occupied block. What happened in 2005 in the Potomac tunnel just before Rosslyn was... this failed. A running train almost tail-ended one stopped in the tunnel, and the one behind *it* also did likewise.  After a lot of work, WMATA engineering thought it was a cable fault, and after repositioning them, could not replicate the problem. As one consequence, the WMATA chief signaling engineer created a test well beyond GRS/Alstom's; his used a test shunt [short between rails] at three places per block [not just one], at each end and in the middle. In 2009, at Ft. Totten, the same problem occurred on a curve. Because of the limited line-of-sight she had, the 112 operator's emergency stop command only reduced the collision from being at >55mph to ~40mph. 112's lead car collapsed, and 9 including the operator died. After weeks of work by WMATA Engineering & NTSB [who was on-site for 68 days], they found that the audio frequency transmitter stage was breaking into parasitic oscillation at the peak of its waveform. That noise was coupling through the common power supply and the metal rack over to the receiver. (Even with the field wiring disconnected, the receiver said "unoccupied.") Further work showed "loss of shunt" cases were occurring hundreds of times *per day* and had been for years. But one cause of these was the GRS parasitic; found on hundreds of other track circuits across the system, all using the same generation of modules, as had Rosslyn. But while that was the core of the technical causes, it was only the first step of NTSB's findings. Over the winter, they had held an evidentiary hearing about the case. During that, the Chairman of the WMATA Board of Directors testified that roughly, safety was not their problem; they just set policy. Perhaps as a result of that statement, the NTSB issued not just 15-odd Recommendations to WMATA (and many to other parties), but one VERY pointed and explicit one directly to the WMATA Board; to step up and take charge of a creating a "safety culture" at the WMATA. As one example of how it is now, the organization is so balkanized that the chief signaling engineer heard only in post-accident interviews that not one of the questioned maintenance technicians had ever heard of his enhanced test. It seemingly had never been promulgated. Since no WMATA Board members were even present for the Recommendations, the NTSB Chairwoman told the (present) current interim General Manager that she and the Members would go with him when he met his Board. In short, while there were technical causes to the accident, the bigger issue is no one in charge seemed concerned about rider or worker safety. WMATA holds the dubious record for unresolved NTSB Recommendations from past fatalities. I wish these were new risks; but they are anything but. One result of this is there is now significant Hill interest in establishing and enforcing federal transit safety regulations. Until now, transit lines such as WMATA & NYCTA have not fallen under FRA regulations. Links: [from <http://ntsb.gov/events/boardmeeting.htm> page] Synopsis <http://ntsb.gov/Publictn/2010/RAR1002.htm> Presentations <http://ntsb.gov/events/2010/Washington-DC-Metro/presentations.htm> Docket, with megabtyes of data. <http://www.ntsb.gov/Dockets/RailRoad/DCA09MR007/default.htm>  WMATA carries ~800,000 person-trips per day. If there is anything to grateful for, it is that the inevitable collision took place in the open on the Red Line, and in the off-peak direction. Fire/Rescue had good access and got accolades from NTSB. If instead, the collision had been within the Potomac tube during crush hour, there could easily have been hundreds killed; each six-car train carries ~1000 people. [Just thinking about such a crash in the tube freaks me.... that's 1500-1800 people in the two trains. The crash would immediately kill say 200+, but the resulting fire would do in another 500+. The tube is NOT sprinklered. DL]
http://www.usatoday.com/travel/flights/2010-08-31-1Acockpits31_ST_N.htm Flaws in flight simulator training helped trigger some of the worst airline accidents in the past decade, according to a USA TODAY analysis of federal accident records. More than half of the 522 fatalities in U.S. airline accidents since 2000 have been linked to problems with simulators, devices that are used nearly universally to train the nation's airline pilots, the records show. Simulator training is credited with saving thousands of lives. But the problem, according to National Transportation Safety Board (NTSB) case files and safety experts, is that in rare but critical instances they can trick pilots into habits that lead to catastrophic mistakes. ["More than half" seems to me overblown. PGN]
The number of pedestrians killed on New South Wales roads in the 2010 calendar year to date, 53, is up 25 per cent over the same period last year although the overall rate of road deaths has dropped. Concern is growing about the "iPod zombie trance" that people get into when listening to mobile music devices. This past weekend there were reports of at least six pedestrians hit by vehicles on the state's roads. Two of them died, one hit by a bus and the other by an ambulance that had its warning lights flashing and its siren sounding. Pedestrian Council of Australia spokesman Harold Scruby is quoted as saying: "'Death by iPod' is a relatively new phenomenon so it may be slow in showing up because it can sometimes be a year between the fatality and the coroner's finding. But we should be asking ourselves why are total road deaths declining while pedestrian fatalities continue to escalate? Maybe listening devices could be part of the explanation.'' The Automobile Association in the UK issued a statement last month expressing concern that people were "increasingly guilty of focusing more on Google Maps while walking the streets than paying attention to the world around them". http://www.smh.com.au/digital-life/mp3s/pedestrian-death-rise-blamed-on-ipods-20100905-14w4d.html Loss of situational awareness poses risks regardless of what people are doing. When they are driving a vehicle or even walking in the streets, consequences can be fatal. http://crave.cnet.co.uk/mobiles/zombie-ipod-pedestrians-endangered-by-mobile-oblivion-says-aa-50000277/#ixzz0yhrj96oa
Surely this is some sort of record. "Public broadcaster ARD's show 'Plusminus' teamed up with the known hacker organization 'Chaos Computer Club' (CCC) to find out how secure the controversial new radio-frequency (RFID) chips were. The report shows how they used the basic new home scanners that will go along with the cards (for use with home computers to process the personal data for official government business) to demonstrate that scammers would have few problems extracting personal information. This includes two fingerprint scans and a new six-digit PIN meant to be used as a digital signature for official government business and beyond." http://yro.slashdot.org/article.pl?sid=10/09/02/1747213:
A lawmaker has uncovered that 1,715 files containing South Korean military secrets, including war plans against North Korea, were stolen from infected Army-issued computers. http://news.softpedia.com/news/Malware-Used-to-Steal-South-Korean-Military-Secrets-153153.shtml
In Nov 2005, three senior aides to Britain's royal family noticed odd things happening on their mobile phones. Messages they had never listened to were somehow appearing in their mailboxes as if heard and saved. Equally peculiar were stories that began appearing about Prince William in one of the country's biggest tabloids, News of the World. https://www.nytimes.com/2010/09/05/magazine/05hacking-t.html?hp
I just received an invitation from Facebook on behalf of somebody I know to join. (I do not, and do not intend to, have a Facebook account.) I haven't heard back from him whether he actually sent it, but it doesn't really matter. Near the end of the email is the line "Other people you may know on Facebook:" followed by eight names and pictures. Seven of the names and pictures are indeed of people I know and except for one correspond with more or less frequently. (That one was of someone I didn't recognize, but I searched my email archives and indeed found one email from him.) As best I can remember, I have never received a Facebook invite from any of them. How could Facebook possibly know with such accuracy who I correspond with on the Internet? A couple of the people I interact with both via email and newsgroups, but several definitely only via email. I hope there's an obvious explanation that someone who uses Facebook would know (like, for instance, if it insisted that you share your address book with it) but whether it's obvious or not it's pretty darn scary.
Robert McMillan, Spammers get the boot with Facebook's new remote logout; The social-networking company is rolling out a new security feature that lets users see which computers and devices are logged into their Facebook accounts. 7 Sep 2010 http://www.itbusiness.ca/it/client/en/home/news.asp?id=59072 Facebook users will soon have a new way of knocking spammers out of legitimate accounts. The social-networking company is rolling out a new security feature that lets users see which computers and devices are logged into their Facebook accounts, and then removing the ones that they don't want to have access. ... Why could a spammer using a stolen account not regularly monitor for other logins under the account and knock them off?
Twitter to log every click on every link in every tweet http://bit.ly/bYtPsp (Google Buzz) "Soon, Twitter will be collecting data on which Twitter users click any links in any Twitter streams. They will also be able to collect IP address info for any user (even non-Twitter users) who click on any link in any Twitter message via the Twitter Web interface." ... [From Network Neutrality Squad. PGN]
here's an item in this weekend's newspaper about the Ford company offering a driving economy check for Ford car owners (in the UK, may apply in other countries too): http://www.ford.co.uk/OwnerServices/FordEconoCheck Basically, they fit a logging device to your car, then you drive round as normal for a week while it records when and where you've driven, then it's mailed back to Ford for analysis. They tell you if you're driving too aggressively or leaving the engine idling too long, etc. and give you tips on how to improve your fuel economy. All well and good, but then the newspaper reviewer advocates this as a great idea for law enforcement—"If the police and insurers ever get this type of information, some serial-offender motorists could be providing damning evidence that might invalidate their policies or land them in court. For the law-abiding majority, EconoCheck will work in your favour... responsible motorists will have data sheets that show they drive sensibly..." http://www.telegraph.co.uk/motoring/columnists/mike-rutherford/7950309/Mr-Money-Big-Brother-is-watching.html Lots of data-protection and security alarm bells ringing here, with the usual RISKS of genuine mistakes, blackmail, and so on. What are the standards for the legal validity of the data, and do innocent people have to prove their innocence?
Robert McMillan, Cisco patches bug that caused partial Internet blackout; A Duke University experiment inadvertently uncovered a bug in Cisco IOS XR. InfoWorld Home, 30 Aug 2010 | IDG News Service http://www.infoworld.com/d/networking/cisco-patches-bug-caused-partial-internet-blackout-811?source=IFWNLE_nlt_sec_2010-08-31 Selected text: In a security advisory released just hours after the incident, Cisco confirmed that an incident disclosed the bug. "An advertisement of an unrecognized but valid BGP attribute resulted in resetting of several BGP neighbors on 27 August 2010. This advertisement was not malicious but inadvertently triggered this vulnerability," Cisco said in its advisory. Duke University assistant professor Xiaowei Yang declined to explain the point of her experiment, but she said that all of the data that her team sent was "100 percent standard compliant." In an interview [on 30 Aug], Zmijewski said that while Cisco's buggy software caused the problems, the Duke team running the experiment should have been more careful. "The days of academics playing with a live network are kind of gone now," he said. "I think it would be foolhardy to try something like this in the future. ... I'm amazed that this happened in the first place."
Randall Stross, A Strong Password Isn't the Strongest Security, *The New York Times*, 4 Sep 2010 "Make your password strong, with a unique jumble of letters, numbers and punctuation marks. But memorize it - never write it down. And, oh yes, change it every few months." These instructions are supposed to protect us. But they don't. Some computer security experts are advancing the heretical thought that passwords might not need to be "strong," or changed constantly. They say onerous requirements for passwords have given us a false sense of protection against potential attacks. In fact, they say, we aren't paying enough attention to more potent threats. Here's one threat to keep you awake at night: Keylogging software, which is deposited on a PC by a virus, records all keystrokes - including the strongest passwords you can concoct - and then sends it surreptitiously to a remote location. ... http://www.nytimes.com/2010/09/05/business/05digi.html [Also noted by Matthew Kruk, who quoted another para: Donald A. Norman, a co-founder of the Nielsen Norman Group, a design consulting firm in Fremont, Calif., makes a similar case. In "When Security Gets in the Way," an essay published last year, he noted the password rules of Northwestern University, where he then taught. It was a daunting list of 15 requirements. He said unreasonable rules can end up rendering a system less secure: users end up writing down passwords and storing them in places that can be readily discovered. "These requirements keep out the good guys without deterring the bad guys," he said. PGN]
I thought I should pass this along because as with almost every type of "software" developed the first ones tend to be crude and the others get more refined, I felt I should report this to make people aware of it. I've seen aggressive anti-virus software, especially in the so-called "free trial" models, most typically in the extent of being aggressive in claiming one's system is horribly infected, with viruses, spyware, malware, and every other piece of worthless or harmful software there is, telling you how terribly bad everything is on your system, and holding its hand out waiting for a tip in the form of demanding you buy the product to fix the multitude of problems it's claimed it found. This rationale of raising the threat level of everything that is on your system that is even in the slightest bit out of line to the level of an unmitigated disaster requiring immediate repair (by purchasing the product, of course) or even worse disaster will result, extends to the newest crop of "fix up" software in which it also looks for things like registry errors, old and unnecessary modules of half-installed, partially-installed or partly-uninstalled applications, in the program declares that every error, mistake or misconfiguration it can find or think of telling you about is about as dangerous to your system as if it was installed by BP or is the software equivalent of its Gulf of Mexico oil spill. Nothing necessarily wrong with this, it reminds me of the days of carnival barkers announcing how you can win a fortune in prizes in what were often rigged games, and newsboys on street corners screaming out lurid headlines to get people to buy the paper, except, of course, the software product always tends to be even more overly alarmist, even to the point in some AV or PC tune-up products of claiming browser cookies for various advertising sites - used for tracking, but otherwise harmless - to be a source of danger. That, and one other issue I'll get to later. I helped out one customer who had a really aggressive product that kept claiming her laptop was full of viruses and spyware and screaming how she really, I mean *really* needed to buy the product before her computer was sold off as part of a botnet to spammers in China or it melted down from the heat generated by the infected software on her computer. I'm not sure if anti-virus software makers think that making their software as obnoxiously irritating as possible is going to make more sales, but there is a "point of diminishing returns" when the anti-virus software becomes so demanding of a tip like an aggressive waiter in a restaurant that you'll choose to go to a different eatery and never go back there again. The software equivalent of this, of course, is the uninstall function. Back in the MSDos days you simply uninstalled a program by deleting its directory, possibly removing a path statement in AUTOEXEC.BAT and that might be it. Now, install and uninstall of the typical Windows application has so many features and options that there are install packages like Installshield and Wise Installer in order to make sure the developer hasn't forgotten anything the app needs and to make sure when the user is uninstalling the application that it's cleanly removed so pieces don't get left behind and unnecessary or now incorrect registry entries are removed. In fact, it's the failure of some uninstallers to either properly do this, do all that is necessary to completely uninstall the program, or do it correctly that PC tune-up programs fix these sort of errors. That is, regular applications tend to provide an uninstall feature and will uninstall themselves when requested. Uninstalling a trial-version of an an anti-virus, software tune-up or similar application is basically an unmitigated disaster for the maker of that product - they didn't get paid for telling you about your problems - and they try very hard to prevent this from happening. A typical application - even a demonstration program - puts an entry in the install/uninstall registry list so you can use "add/remove programs" in Control Panel (at least for versions of Windows through XP; I'm not sure what the program removal tool is called on Vista as I really don't understand it much and I've never used Windows 7). A number of these "anti virus" or tune-up applications won't put an entry in add/remove programs. Obviously, they do not want you to shut up their screaming banshee applications without paying for them, they want you to buy them off! So going back to the lady with the aggressive anti-virus on her laptop, it didn't have an uninstall but I figured out where it was and basically ripped it out by the throat, deleted its directory and related files and if anything showed up as missing during a reboot, remove references to those files too. That was about two years ago. So my sister calls me last week because for the first time now she has an "anti-virus" program that is warning her about a bunch of problems on her system, and even refusing to let her install the Internet telephony application Skype. Basically this program is telling her that almost everything is infected and she needs to fix it (by buying the full version of this "anti-virus" program, natch.) I'll explain why I put "anti-virus" in quotes in a moment. I take a visit over to her place and find some "anti-virus" product whose name I can't remember exactly, but I think it was something like System Protector, or something like that. She doesn't even know how this program got on her computer. Basically it was a one-trick pony, telling how just about everything on the system was infected with a virus that was attempting to "send your credit card numbers over the Internet" to someone in Russia or one of the various countries well known for credit card fraud. It's blocking just about everything as being infected. My sister called me because she was having trouble being able to use her computer - from all the warning messages of this so-called "anti-virus" software - or being able to use the Internet. So it's going along and telling me about all the stuff on the system that is horribly infected with viruses that are trying to send my credit card information over the Internet to who knows where and how it can stop this if only I "upgrade" to the paid version, I'm trying to figure out how to shut it up so we can actually use the computer. The third most significant thing that made me suspicious was, to the extent I could find the equivalent of add-remove programs in Windows Vista, this program is, of course, not present; but nothing new here, this happens often with many of these kinds of programs. The second thing was its splash/main screen which simply had its name, it had no other branding than then name of the program (which again I'm calling System Protector because I can't remember its name). Every other real anti-virus application has a company brand in addition to the product name, e.g. Symantec, McAfee, Kapersky, etc. Nothing with this one. But what got me most suspicious was its claims that essentially everything is trying to export credit card numbers to unsavory characters in lawless parts of the globe. Not everything is doing this and claiming it is clearly the worst kind of alarmist hysteria. My sister runs behind a firewall, there couldn't be that many things infected as System Protector claimed. It started when I can't access CMD.EXE, because "it's been infected with a virus that is trying to transmit credit card numbers" etc. and won't allow me to run it, but I can fix this by purchasing the full version, etc. It even prevents Adobe Flash from trying to install an upgrade, so maybe it's not all bad. Of course, I have no idea why a download from Adobe's website has a virus trying to send credit card numbers, but the program must know what it's talking about, right? I basically can't use the Run command, because everything I'd try to run is, you guessed it, trying to forward credit card numbers to Afghanistan. When I could not run Regedt32 because the program thought it was forwarding my Visa and MasterCard information to Osama Bin Laden, I came to the realization that basically this wasn't an anti-virus program, it was a trojan faking an anti-virus and holding your computer for ransom until you buy the full version of their so-called "anti-virus" whereupon you'll be lulled into a false sense of security since it will now stop telling you you're infected. It's the high-tech version of selling counterfeit prescriptions in which you think you're getting a necessary medication but you're actually buying something that is, at best a do-nothing sugar pill or possibly something full of harmful chemicals. It is basically blocking anything you could use to disable, remove or shut it down without paying for its useless "protection." After some effort I realize I can stop it by rebooting into Safe Mode. Now while I can't find what directory "System Protector" is stored in, I can run Regedt32 and discover that there is an entry in the Run Once and Run entries consisting of a name in Chinese! So the so-called "anti-virus" has fixed itself to be restarted any time the computer is rebooted (other than in Safe Mode, fortunately). So while I couldn't remove the application because I couldn't figure out how to translate the Chinese characters into a directory, I was able to evacuate it from Run Once or Run, and when the computer was restarted, the "anti-virus" was gone; I had stopped its execution. One of the things I referred to earlier was the old days of newspaper vendors hawking lurid headlines to get you to buy their paper. But at least they weren't blocking your way and refusing to let you leave unless you buy a copy. This is clearly extortion disguised as legitimate software protection. The obvious risks are first, that someone will believe this and pay them, maybe becoming a victim of who-knows-what; if they can fool you into thinking it's a legitimate anti-virus tool, it might have any number of nice features including actually being a botnet manager or who knows what. If they ripped you off one way, no reason they can't find a way to rip you off another way, and maybe another. Second, since it surely isn't going to find any real viruses, it will give someone a fake sense of security by thinking they're protected. As I said, this one was crude and way too obvious, but I suspect that there will be other programs similar to this in the future which will be more refined and harder to spot as fakes, and I felt people should be aware of this sort of thing looming on the horizon. And people used to wonder why I never use Internet Explorer and always have a hardware firewall in place. I don't use anti-virus software and have never gotten an infection. While maybe I'm just lucky or careful, it's been over 15 years of using the Internet nearly every day and the number of times I've gotten infected on my computers has been and is still zero.
[This may be a timed-out e-mail address. PGN] I found 4 security problems at on a bank's website. There were 3 security problems that came up (due to the flawed mechanism that the bank has for reporting security problems). I coincidentally found a 4th problem that may also be a security problem. Here are the problems: Initial security problem: 1. When logged on to the banking site and trying to send a message to the bank it is possible for the session for the message sending to timeout. That browser window says that the user timed out and he has to log in again to send the message. While not actually saying that the user was logged out of his main banking session, most users would assume that they were and it was safe to leave the terminal. However, in fact the user may not have been logged out of the main banking session and can actually do banking or send a message without logging in again. Problems found while trying to report the problem to the bank: 2. The message center messages can't be used to report security problems. 3. When you use the phone to report a security problem, you are asked to prove you are a customer by giving your account number, even though you might not want to give your account number because of the security problem. Also the human response that says you have to call a phone number to report the security problem should have contained a token that allows someone to confirm over the phone that he is a customer, AS SHOWN BY him getting the token which requires the user to enter his banking information. 4. There is no way to report a security problem if you are not a customer. I also found that the bank's message system incorrectly processes the included original message in reply to threads. In particular, each apostrophe seems to get changed into two apostrophes, although there might be another factor involve in this doubling. I consider this a security problem since it indicates that text is being process someplace, rather than merely being blindly included with an include string stuck at the start of each line. This type of processing often turns out to indicate a security hole whereby text winds up getting executed with bad results.
Tim Jones, *Wall Street Journal*, 4 Aug 2010 In addition to the comments in the article, the technique usually makes it possible for the 3rd, 4th, and 5th parties involved to violate the 1st (user's) party's privacy, the privacy agreement that the 2nd party (the site) has made with the 1st party, and the privacy agreements that the actual advertisers (4th or 5th party or beyond) have made with the 1st party. Underlying principle: Of course this information personally identifies the 1st party - that is why the ad has been placed here in the first place.
Please report problems with the web pages to the maintainer