At this time of year we enjoy the twice annual collection of stories about problems caused by time zone adjustments. DST is a cunning way of getting people to adjust their habits to make better use of sunlight when it is available. We know from the turbulent history of DST in the USA that people will not make this adjustment without external influence, or if they do they will not do so with consistent start and end dates or indeed any regard for the inconvenience of those around them. (See David Prerau's book, "Saving the Daylight".) So DST is beneficial provided it is applied consistently over a reasonably large area. However it is a crude and arbitrary mechanism. It offends those who think time should be a matter of natural philosophy, not of politics. It is a great inconvenience to us technologists when the politicians cannot stop themselves from messing around with the schedule. It causes many problems when the clocks suddenly jump by an hour twice a year. I believe there is a way to enjoy the benefits of DST while avoiding these drawbacks. The essential idea is that our clocks should be set using sunrise as a benchmark instead of noon. This is an entirely scientific way of adjusting our clocks (and therefore our habits) to seasonal conditions, so it is immune to political fiddling. Our clocks would run fast by about a minute a day in the spring, and slow by a minute or two a day in the autumn, so there would be no unpleasant disruptions to our sleep. If we forget to make an adjustment we won't be embarrassingly early or late. It is obviously not sensible for clocks in Land's End and John O'Groats to tell different times just because of their differing latitudes. Therefore, just as we use standard longitudes to define our time zones, we would use standard latitudes to define sunrise time. Let us use the time of sunrise at the tropic of cancer, 23.44 degrees north, as our standard. The difference between this time and that latitude's latest sunrise, 06:44, gives us an offset to add to our zone's standard time. This adjustment varies smoothly between nothing in January and an hour and a half in June, giving us even more evening sunlight to enjoy. Southern countries would use the same mechanism, but with the tropic of capricorn as their standard latitude. Some will argue that it is inconvenient to adjust one's watch every day for most of the year. We were happy enough to do so with mechanical watches in the past, so I don't think this is a big deal, and lazy people can probably get away with adjusting theirs once a week. I also see it as an opportunity for innovative new intelligent clocks and watches. There may be slightly more difficulty checking relative times when communicating between northern and southern sunrise time zones, but the time difference tables will only be about 40 times larger. It is also a great way for geophysicists to remain involved in timekeeping after leap seconds are abolished. I recommend this proposal to you, and hope that it is as successful as William Willett's idea one hundred years ago. f.anthony.n.finch <email@example.com> http://dotat.at/
The April 2008 issue of the *Communications of the ACM* includes an important Inside Risks article by Lauren Weinstein. (It is of course subject to CACM copyright, so I won't reproduce his article here, but suggest that it is worth reading.) It is online on my Inside Risks website: http://www.csl.sri.com/neumann/insiderisks08.html#214
The last publication of the Chaos Computer Club (CCC) has published a fingerprint of the Interior Minister Wolfgang Schäuble (quoting the oft-heard mantra "if you have nothing to hide you should have nothing to fear"), together with a tongue in cheek "collection album" page where readers can fill in fingerprints of other ministers if they manage to collect them. http://www.ccc.de/updates/2008/schaubles-finger (sorry, only in German). The CCC didn't stop there: for good measure they also repeat their 2004 guide in both English and German on how to lift fingerprints and use them as your own, complete with links to videos of the process and how it has been used to defeat a pay-by-fingerprint system of a German supermarket chain. http://www.ccc.de/biometrie/fingerabdruck_kopieren?language=de (German) http://www.ccc.de/biometrie/fingerabdruck_kopieren?language=en (English) The usual "we'll sue you" noises are already being heard, which highlights interesting questions about the fingerprints you leave behind..
The following cartoon makes an amusing observation about the recently increased mismatch between European and American DST schedules. http://www.telegraph.co.uk/money/graphics/2008/03/28/calex28.gif f.anthony.n.finch <firstname.lastname@example.org> http://dotat.at/
My view has been and always will be: Australia & NZ should totally abandon Daylight Savings Time (DST). DST has no place in Australasia because most of Australia and NZ are Semitropical or Temperate—with the corresponding reduced variation in sunrise and sunset times. The only region of Australasia that may even be nominally affected by this change are the NZ provinces South of Canterbury (where Christchurch is, South Island). As part of the this region's attempts to reduce its carbon (CO2) output, a policy of reasonable workplace scheduling needs to be instated. With the abolition of Australia's "Work Choices" and some minor tweaks to NZ employment contracts laws—this can be done without disenfranchising anyone. As a matter of state policy, the Australia & NZ "Fee Trade Agreement" (FTA) and the "Uniform Commercial Code" (UCC) needs to be amended to abolish DST, as it creates a "NONUNIFORM competitive environment." Using DST is probably more responsible for the loss of global competitiveness in Australasia, as it creates totally unnecessary work in the commercial and governmental sectors—and needlessly endangers people's lives. I hope the new Rudd government issues a Y2036/Y2038 compliance law that forces the Federal and State governments to Audit their systems and gradually impose compliance benchmarks as time goes on. The Unix/POSIX time problem will negatively impact Australia's (and NZ's) global competitiveness if it is allowed to remain unfixed. Max Power, CEO, Power Broadcasting HireMe.geek.nz Mini-Y2K fears over Aussie daylight saving change // By ASHER MOSES - SMH | Friday, 28 March 2008 http://www.stuff.co.nz/4454030a28.html The decision to extend daylight saving in south-eastern Australia could create a mini-Y2K by putting the internal clocks on computers, smartphones and corporate servers out of sync. From this year on, daylight saving in NSW, Victoria, ACT, Tasmania and South Australia will end a week later than usual on the first Sunday in April and, with the exception of Tasmania, recommence three weeks earlier on the first Sunday in October. The change was intended to harmonise daylight saving dates across the country and give Australians more daylight hours, which in turn benefits the environment by reducing evening electricity use. Many electronic devices with internal clocks are set to adjust automatically for daylight saving but, as a result of the recent date changes, the adjustments this year will be incorrect. The fallout for regular consumers could include missed meetings or appointments, but corporations face bigger headaches as their internal servers, fleets of BlackBerry devices and automated systems such as payroll, stock trading and manufacturing are operating under the old daylight saving regime. Clocks must therefore be adjusted manually or via software updates from the device makers. A similar issue occurred in the United States last year when daylight saving was changed to kick in three weeks earlier and end a week later. At the time The New York Times reported it would cost public companies $US350 million to make computer fixes to deal with the changes. Microsoft has issued an advisory to users of its Windows, Outlook and Windows Mobile products recommending they download an update from microsoft.com.au that will synchronise computer clocks with the daylight saving changes. "The synchronisation [issue] is not exclusive to Microsoft products. It affects all devices that update automatically according to the old daylight saving schedule," Microsoft's customer and partner experience director, Hugh Jones, said. IDC analyst Liam Gunson said widespread problems could occur if people were not made aware of the issue and did not take action to fix it. He said the same problems were predicted in New Zealand last year when daylight saving changes were made but no serious problems eventuated. "It was really just a matter of education and people knowing that they need to download a certain patch or look at their IT systems and it appears that most people did," he said. The issue has been likened to the Y2K or millennium bug, albeit on a far smaller scale and with less serious consequences. Y2K caused chaos leading into the new millennium as it was feared computer systems, which stored years as only two digits, would be unable to recognise dates from 2000 onwards. Governments spent hundreds of billions of dollars working to fix the problem, with computer engineers predicting doomsday scenarios such as that critical finance and electricity industries would stop operating and planes would fall out of the sky. However, when the year 2000 finally arrived, there were no major computer disasters. There is debate over whether this was a result of the immense preparation for Y2K or people overstating the seriousness of the problem.
*The Village Voice* reports that the New York City Police Department's "CompStat" report for the 9th Precinct shows zero homicides in 2008. In spite of Tina Negron having been murdered in an East Village supermarket on February 29, 2008: You have to go to the fine print - an asterisk at the bottom of the stats - to get what's kind of an explanation: "Crime figures for February 29, 2008 ... were excluded to ensure accurate comparisons." Negron wasn't the only victim who was victimized again by the stats. A total of 248 felonies, including two murders, occurred citywide on February 29. But they were excluded from the CompStat analysis - the NYPD's method of tracking seven "major" crime categories (murder, rape, robbery, felonious assault, burglary, car theft, and grand larceny). [...] The NYPD press office's top CompStat guru didn't return several phone calls from the Voice. But according to published reports in 2004, the NYPD stopped counting Leap Day statistics in 2000. Attributing the reasons to an unnamed police spokesman, a Daily News story explained that Leap Day is withheld from CompStat because "adding the extra day ... could show an unreliable increase in crime in comparison with the prior weeks and months and cause changes in deployment when it is not really necessary." Full story at: http://www.villagevoice.com/news/0812,The-NYPD-Ignores-Leap-Day-Crimes,381244,2.html The cooked statistics from the NYPD for the 9th Precinct are viewable here: http://nyc.gov/html/nypd/downloads/pdf/crime_statistics/cs009pct.pdf [note that the footnote that crime stats for Leap Day were excluded does not appear in that PDF, but it does appear on other CompStat reports at the same web site] See also my post in RISKS-13.69 describing how the NYPD played computer games with a performance metric in their 911 dispatch system, and RISKS-24.28 on much more blatant (and unauthorized) rigging of the crime statistics in a different precinct by a high-ranking cop who wanted to improve his numbers. Crime statistics have been used as political bludgeons for years in NYC and it's not surprising that the NYPD takes every step possible to avoid looking bad. I wonder what crimes happened on February 29, 2000, that prompted that policy change in the first place? [Also noted by Danny Burstein, who noted that other big cities (such as LAPD) include leap-day numbers. PGN]
More flights canceled as Heathrow remains in chaos By Alan Cowell The New York Times Friday, March 28, 2008 British Airways canceled dozens of flights at Heathrow's glittery new Terminal 5 on Friday as its staff struggled for the second day with state-of-the-art technology that was supposed to hasten check-in procedures and make flying a pleasure. The hitches since the terminal opened to passengers on Thursday were "definitely not British Airways' finest hour," the airline's chief executive, Willie Walsh, said as he offered a personal, public apology for disrupting the travel plans of thousands of people. British Airways canceled almost 70 flights on Thursday, after a day of delays caused by baggage handling problems. On what was supposed to be the first full day of operations at Terminal 5, many flights took off with their holds empty, carrying passengers with just cabin baggage. Some passengers slept overnight in the steel-and-glass terminal - reviving precisely those images of delay and decline in British aviation that British Airways said it would banish with the opening of the new terminal. As a result, Walsh said, about 36 flights out of Terminal 5 - mainly short-haul and domestic - were canceled in advance Friday to ease pressure on staff members dealing with unfamiliar procedures and systems. Walsh said there had been "problems in the car parks, airport areas, computer glitches and the baggage system." About the prospects for the weekend, he said Friday: "I would expect some disruption tomorrow, but I think it will become better as we become accustomed to the building and the quirks of the systems." Travelers arriving early Friday confronted what one traveler, Tony Pascoe, 35, called chaos as they stood in line for several hours only to be told their flight had been canceled. "It was chaotic," he told Britain's Press Association, "Everyone who had been queuing were annoyed and a lot of jostling and arguing started. Then the desk just crashed so everyone stood there. "It is diabolical. I am a frequent traveler and this is the worst experience ever - it is absolutely shocking." "This is a public relations disaster at a time when London and the U.K. are positioning themselves as global players," said David Frost, director general of the British Chambers of Commerce. "We can only hope that this will provide a wake-up call as we gear ourselves up to host the Olympics in 2012." Heathrow is one of the world's busiest airports, handling about 67 million passengers a year. The new terminal - reserved exclusively for use by British Airways - was designed to counter the airport's image as an unpleasant place for travelers. The building cost about $8.7 billion and has 10 miles of baggage-conveyor belts supposed to carry up to 12,000 items of luggage an hour. But the baggage system has been at the heart of the start-up problems. Other airlines, excluded from Terminal 5, took some delight in claiming to pick up business from British Airways as travelers switched to carriers operating out of Heathrow's older terminals. And a private aviation company, Netjets, said in a statement that the number of people seeking private business flights had risen by 88 percent over a 24-hour period as "travelers sought to bypass the chaos of the opening of Terminal 5 at Heathrow." http://www.iht.com/articles/2008/03/28/europe/heathrow.php Archives: http://www.listbox.com/member/archive/247/=now
I assume other comp.risks contributors will by now have provided the details and the background regarding the problems of Heathrow's terminal 5: the parking sign snags, the baggage processing backlog, the canceled flights, and the resulting chaos. A related interesting angle is an email that British Airways circulated to its customers on the day of the terminal's opening. Here are some notable excerpts, as highlighted by a colleague who brought this to my attention: - - - - Dear Mr [...], Five and a half years ago the building of our new home began in our most visionary project to date. Today we opened the doors. There is no more waiting... Terminal 5 welcomes you. *At Terminal 5 everything has been streamlined and designed to make your journey through the terminal calm and relaxed.* And this morning we saw all the planning fall into place. The next time you fly in to, or on from Terminal 5, *you'll experience for yourself how all the planning and careful design has fallen into place.* The arrivals Gates are conveniently located to minimise your walk from the plane and if you're transferring to another flight, Flight Connections is so smooth, you'll be through in 20 minutes. *A state-of-the-art baggage system*, a shopping concourse that rivals London's West End, and an array of tempting restaurants, bars and cafes to choose from, you'll discover nothing has been overlooked to ensure *your time at Terminal 5 is spent in a most relaxing and enjoyable way.* [...] In this case the risk is that the making of grandiose claims about yet-to-be-established performance can easily backfire. Diomidis Spinellis - Athens University of Economics and Business http://www.dmst.aueb.gr/dds
re: http://catless.ncl.ac.uk/php/risks/search.php?query=gps I note, after searching this RISKS database of items on "GPS", that a considerable number of observations from your writers re GPS errors are actually errors in the mapping data bases that are used in navigation system applications (e.g., automotive navigation), rather than a GPS positioning error due to signal errors per se. This distinction may not be interesting when you are lost in your car, but it is critical in other applications. GPS position estimates have inherent errors (generally of a couple of meters in "open sky" circumstances, but possibly 100s of meters on some occasions due to "non-line-of sight multipath error" in especially built-up urban areas. Some GPS-Auto-Nav users will have noted temporary errors such as their position being displayed on the wrong road. The difficulty is more subtle than writers surmise. There are indeed errors in the maps being used. Even if a map is correct when installed in your device, roads change. But at any one moment how can you be sure an error is in the positioning estimate or on the map. You really need to rely on signage if it is available. But worse than all this is that we are on the cusp of deploying GPS-based road-tolling systems, the majority of which will depend on map-matching algorithms to determine which road you are on or which "cordon" you are in to calculate a charge. These tolling systems will be subject to error for the same fundamental two reasons signal errors and map errors. The risk here is that tens of companies are building and tens of municipalities and tens of counties are considering investing in GPS-tolling systems that will critically rely on map-matching. Considering that the very first such system (Germany) cost far in excess of Euro 10^9, these companies, cities and countries are about to put many, many billions at risk. Any decent lawyer could cobble together a class action suit to defeat charges based on map-matching. They only need your collection of emails to show negligent system design. Bern Grush, Chief Scientist | skymetercorp.com desk +1 416 673 8406 | cell +1 647 218 8600
This isn't actually a design flaw or oversight. Naval vessels (like every other ocean-going ships) are equipped with surface search radar, but naval vessels often don't use it. RADAR emissions can be detected at twice the distance they can 'see', so a warship running it's surface search RADAR is both broadcasting it's position and telling everyone how far away they can stay and not be detected. That's often not the most useful thing a warship could do. The real failure here was undoubtedly much more complex than simply not running the RADAR though. The underway watch team charged with safe operation of the ship (i.e. those actually involved in navigation and maneuvering) on a military vessel usually includes a couple of dozen people, including several equipped with nothing more sophisticated than binoculars and a sound-powered phone. That all of them missed seeing the boat until they hit it speaks less of electronic failures and more of some kind of systemic personnel issue. Rick Damiani, Applications Engineer, The Paton Group, California: (310)429-7095
The "trailing zeros" bug Rick Damiani wrote about in RISKS 25.09 reminded me of a similar, but fortunately far less intrusive, problem a friend of mine had with his ADSL connection. I had recommended the ISP I had recently begun using, and he'd happily signed up and got his modem and router configured and working perfectly... well, mostly perfectly. A few web sites, without any apparent relation, just wouldn't work when he went to them with his new DSL account. Switching back to the old account, everything was fine. (And I'd thought PPPoE could never have a benefit.) Since I'd recommended the ISP, I was on the hook here, especially since my connection had been, and continues to be, quite reliable. So I did the usual pings and traceroutes and didn't notice anything other than the usual "ICMP is scary" lossage. No two of the failing web sites seemed to be network-ologically related, so it didn't look like a particular carrier having issues with that ISP... and, anyway, I could get to all of them--via the same hops. In desperation, we went into his router's set-up. It didn't _feel_ like a Path MTU discovery problem, but I was out of ideas. Then I noticed the IP address of his modem: x.y.z.0/32. A perfectly legitimate host address for a point-to-point connection. So we called up the ISP's support desk, and told the guy there what was happening and my suspicion about the "trailing 0" being a problem. It wasn't _wrong_, but it was the only thing odd I could see. The guy at the ISP agreed, right down to the "it's not wrong but it's unusual" feeling, and assigned a new IP with a non-zero final octet to my friend. Sure enough, all the missing web sites turned up. My guess was that some providers were dumping packets purporting to be from a /24 network address, making the assumption that an all-zeroes final octet must mean the packet is spoofed. Which is fine for /24 all the way up to /31. But for anything else, you're at RISK of having a legitimate host address junked. /24 is common. Really, really common. But we all know the RISKs that arise when we treat "common" as if it was "only". You can't tell what my address structure is; even before CIDR, I was regularly working in subnetted class A space, and our netmasks never left the building. (Either that, or someone had heard the old saw that "auditors reject any line item that ends in 5 or 0.")
In RISKS-25.09, Mark Brader wrote a submission with the subject: "Hoax on Craiglist causes duped victims to steal property." A demonstration of how making the "long story short" changes the story completely. [PGN-ed and oversimplified; don't blame Mark.] The victim was not unsuspecting when he returned home. He had received a phonecall while away from home from someone about the horse, which was in much better shape than it should have been had it been abandoned. While driving home, he passed several people with truckloads of property he knew was his. When stopped and told they had his property, they ignored him. When he arrived home, he found more people, some of whom showed him a printout of the craigslist entry as proof that they could steal his property, and many of them drove off with more of his stuff, after being told they were stealing. There were no "duped victims". The victim cannot, by definition, steal his own property. Those who stole were dupes, but they aren't the victims here in any reasonable sense of the word. The people who got the property profited. The local sheriff has already gone on record as saying that those who took the property face criminal charges if caught, but have been given an opportunity to return what they took with no questions asked. Let's not allow technology cloud the ethics and results. Sometimes dupes are the victims, as in 419 scams, but here the victim was the fellow whose property was stolen. Those who were presented with a "too good to be true" opportunity this time are the thieves, and could have prevented a lot of damage had they simply called the fellow whose stuff they wanted to take to make sure.
Please report problems with the web pages to the maintainer