Geek stuff – Page 122 – a sibilant intake of breath

Regression to the mean

All manner of diets, supplements, and vitamins compete for customers and adherents. Given that we are creatures of biochemistry, it is plausible that such chemicals will have effects on human health. Unfortunately, they do not act alone, but rather within a complex web of interactions: genes, environmental effects, physiological changes, etc, etc, etc. This makes it exceedingly difficult to isolate and prove the effect of any particular substance, especially given that the effect in different people, or the same person at different times, may differ.

The phenomenon of regression to the mean is especially confounding when it comes to individuals. We do not generally change our regimen of diet or supplements unless we perceive something to be wrong. We take vitamins when we feel ill, and analgesics when we have a headache. Given that both illness and headaches tend to rise to a peak and then taper off naturally, virtually any action we take in response will precede an improvement. It doesn’t matter if you spend a pile of money on remedies, spend your time praying, or simply sit still and wait for improvement. Regression to the mean is the product of basic statistical mathematics and isn’t caused by anything chemical or biological. It is tautological to say that things are normally like the mean and that, most of the time, situations far away from the mean will be replaced by those closer to it. While it is true that we can do things likely to shorten or lengthen the period of illness or discomfort, it is virtually impossible for an individual to know whether such an effect has occurred. Did taking those vitamins shorten or lengthen the cold? Did it have no effect? What about those glasses of red wine?

The only sensible course of action is to essentially disregard our own experiences, except in such cases where there is both a reasonably large body of evidence (ideally in the form of a large number of double-blind and controlled trials) and there is a plausible explanation for the method of action. Failure to employ such checks against hasty false reasoning leave us vulnerable to the pernicious human tendency to see causal relationships everywhere, without the scepticism that is critical in separating conjecture from an investigated hypothesis.

A query for any lurking physicists

I was having a conversation this afternoon about the Tunguska event: a huge explosion that occurred in Russia in 1908. I had always heard that it was caused by a meteor impact, though apparently some other explanations have also been considered. One that I just heard about is the possibility that it was caused by a collision between the earth and some antimatter.

Wikipedia suggests that this explanation isn’t credible, but it does leave me wondering: in the event of a collision between a particle of matter and a particle of antimatter, where the two particles are traveling at different velocities in opposite directions, how does the net momentum of the two translate in their annihilation? If antimatter did hit the earth, it would start to strike particles of matter in the upper atmosphere, the particle pairs would annihilate one another, and energy would be released. Would all that happen before the antimatter hit the ground? Presumably, it would strike its mass in antimatter before then. If so, what would the effect on the surface of the planet be?

Conservation of energy dictates that the kinetic energy in the faster particle would need to go somewhere. Presumably, it would manifest in the production of more energy during the annihilation event. As such, I suspect the antimatter clump would get blasted apart in the upper atmosphere and produce some kind of horrible shower of radiation, though nothing in the way of direct physical debris.

Mifare RFID tags reverse engineered

I have written before about security weaknesses in pin-and-tumbler mechanical locks. I suggested that electronic token based systems have a greater capacity to be secure, since they do not rely upon mechanical parts that can be manipulated.

Of course, poorly designed electronic systems can also be breached easily. That was demonstrated in September, in relation to the KeeLoq system used for keyless entry in many cars. Now, another brand (Mifare) of RFID tags have been reverse engineered and found wanting. As is usually the case on matters of physical security, I saw this story first on blackbag.

Book project: month two

Entering the second month of our reading agreement, neither Emily nor I has finished the first book. Allowances can be made, however, for the fact that December ended with holidays and my visit. I am aiming to finish Love and Hydrogen in the next few days and move on to Nabokov’s Laughter in the Dark soon after. No doubt, she will be through the more hefty A History of Warfare before too long.

Despite my nervousness about assigning a second military-themed book in a row for Emily, I have given her Ender’s Game for January. It is quite a compelling read and it serves our original purpose of sharing books that have meant a lot to us and influenced us somehow. Not only have I read this Orson Scott Card novel dozens of times, but it was a pretty important aspect of the collective knowledge of some of my closest friends in high school.

Prior posts:

An orderly transfer of power

When I saw a camera markedly superior to the one I have been using for the last two years on sale for about $150, including a 2GB memory card, it seemed that the time to upgrade had arrived. I was drawn to the Canon Powershot A570 mostly because of the image stabilization, which allows sharper photos in lower light. It is also nice that it has ISO ratings going up to 1600 – compared with 400 on my old A510. It remains to be seen how the graininess of the two cameras compares at fast speeds. The controls on the new camera are nearly the same as the old, though it will take a while for them to become as utterly intuitive as the A510 was after its years of valued service.

What surprised me most about the A570 is how pleasantly quick it is. The time lag from pressing the shutter to taking a photo is much shorter. All sorts of other camera operations are faster too; transferring photos to my computer is about three times faster. The Digic III processor is probably responsible for most of that. As you can see from the two linked images, the A570 also seems to blow out highlights less than the A510. Those frequent white patches were one of the most substantial failings of a camera that is excellent overall.

The old camera remains perfectly serviceable. Virtually every picture posted on this blog has been taken with it. It will probably be available at low cost to a friend who will use it well.

[1 January 2008] The camera has passed to Emily Horn. May she use it well.

Fibre jam

Some people are predicting that 2008 will be the year when the internet slows down. The cause is expected to be massive amounts of video traffic, partially driven by social networking sites. All those voice-over-internet phone calls will naturally add to the flow of packets that need to be routed around the world.

All this makes me wonder whether it might be better to allow a bit more discrimination in routing. Sending and receiving video is fun, but rarely essential. Having the more prosaic uses of the internet suffer unduly because of such things seems improper. I could probably surf text all day using the bandwidth required to watch a few minutes of television online.

I don’t really know enough about internet architecture to be able to say whether such filtering could be accomplished, whether it would be cheated very easily, or whether it would cause additional problems. That said, you can certainly expect such questions to get asked more often if predictions of slowly loading websites and jerky video calls prove correct.

Hacking Nintendo hardware

If you want proof that allowing people to tinker around with commercial hardware can produce cool results, take a look at this video. It shows a Wiimote mod done by Johnny Chung Lee that allows head tracking for the simulation of a three dimensional environment on a flat screen.

I wonder when we will see the first game that uses this approach. It could probably have some more serious applications, as well.

The sheer hackability of Ninendo hardware is making me seriously consider getting a DS. My brother has a card that lets you store dozens of games on a micro-SD card, as well as run homebrew applications.

[Update: 5 January 2008] I finally managed to track down a DS. I have put in an order for a device that will allow it to play ROMs from a micro-SD card, as well as run a web browser and other goodies.

[Update: 9 January 2008] For a very long time, I thought no Zelda game could top A Link to the Past. While it is too soon to know for sure, I can say that The Legend of Zelda: Phantom Hourglass is in the running. Playing with just the touch screen works surprisingly well. The DS is the most entertaining system I have used since the original Playstation.

[Update: 11 January 2009] I received an M3 DS Real Nintendo DS storage device in the mail today from EchoStore.com. Unfortunately, it doesn’t seem to have been properly manufactured and only functions intermittently. A number of web forums discuss making them a bit thicker by taping on pieces of paper. While this does seem to improve the success rate, it isn’t perfect and I would prefer one that works properly. As such, it seems destined to be returned to the manufacturer for exchange.

Copper indium gallium selenide solar cells

Nanosolar, a company supported by Larry Page and Sergey Brin (the founders of Google), has announced that it will be selling thin-film solar cells profitably for $1 a watt. Apparently, the cells are printed with copper indium gallium selenide – an alternative to silicon. Cells based on the material can convert solar radiation to electricity with 19.5% efficiency. In theory, this material can applied to foil, plastic, glass or cement – producing electricity generating surfaces. It can also be made into more conventional panels of the sort Nanosolar is starting to sell.

In the 1950s, solar cells cost about $200 per watt. By 2004 they were down to $2.70. Further reductions could make solar power cost competitive with fossil fuels, potentially even in the absence of carbon pricing. Combined with either better storage (to moderate light/dark and sunny/cloudy cycles locally) or better inter-regional transmission (the sun is always shining somewhere), such cells could eventually make a big difference in the overall energy balance. Solar has been discussed here previously.

Impressions of Leopard

When I upgraded by Mac to Tiger (Mac OS 10.4), I wrote a bit about it. Now, it is Leopard‘s (10.5) turn.

The big new features are Time Machine (backup utility) and Spaces (multiple desktops). There are also incremental improvements to lots of prior features: Dasboard, the Finder (Coverflow added), Mail, Preview, Quick Look, Safari, and Spotlight. Time Machine is a good idea and seems to work well. The graphics strike me as a bit overdone. A simpler interface would use fewer system resources and might well be easier to use. That said, making backups a lot easier is a very welcome move. Startlingly few people have even a single comprehensive backup of all their data, much less the kind of rolling, iterated backup Time Machine produces. It also manages to do so in a far quicker and less obtrusive way than free options like Carbon Copy Cloner. Spaces is too clumsy to be of any use on my iBook. The F1-F12 keys already have too many demands made of them by screen and volume controls, Expose, and Dashboard. As such, I only briefly enabled this feature before rejecting it as essentially unusable. The new firewall also seems more confusing and troublesome than the old one.

Is Leopard worth the $100? Some of the little improvements are certainly nice. I like the Stacks feature that has been added, as well as the way the wireless network icon in the menu bar now shows which networks require passwords. That said, the improvements are relatively minor. I would not run out to buy this upgrade again. That will situation will probably change when common pieces of software begin to require it and developers begin to make better use of the new under-the-hood features.

One of the best things about a new operating system release is that it gives you the chance to prune things down. With Tiger, I used the option that simply replaced the operating system, leaving applications and data intact. That worked brilliantly but couldn’t be done this time, since it was a complete operating system failure that provided the immediate impetus for the upgrade. Backing up data, erasing everything, and starting over has left the computer running noticeably quicker. That is especially welcome on a system that is getting on in years and having increasing difficulty running applications smoothly.

I have taken this opportunity to abandon some more Microsoft software. Previously, I used Entourage (the Mac version of Outlook) because it was compatible with all the emails from my old PC. On the PC, I chose it originally because it would sync easily with my Palm Pilot (which has spent about four years collecting dust in a box in Vancouver). Switching to Mail and iCal was pretty straightforward, since I decided not to transfer over my old emails and to simply synchronize iCal with my Google Calendar. I don’t really like the Mail interface very much, but it does integrate better into the OS than Entourage did, as well as using fewer system resources.

All told, the Mac does feel as though it got a new lease on life. It will need to endure until I can justify redirecting sufficient funds from student loan repayment to buy a shiny MacBook.

Salmon farming and sea lice

Recent work by Martin Krkosek of the University of Alberta has demonstrated strong links between the practice of salmon aquaculture and the incidence of sea lice infestations that threaten wild populations. One study used mathematically coupled datasets on the transmission of sea lice (Lepeophtheirus salmonis) on migratory pink (Oncorhynchus gorbuscha) and chum (Oncorhynchus keta) salmon. They concluded that:

Farm-origin lice induced 9–95% mortality in several sympatric wild juvenile pink and chum salmon populations. The epizootics arise through a mechanism that is new to our understanding of emerging infectious diseases: fish farms undermine a functional role of host migration in protecting juvenile hosts from parasites associated with adult hosts. Although the migratory life cycles of Pacific salmon naturally separate adults from juveniles, fish farms provide L. salmonis novel access to juvenile hosts, in this case raising infection rates for at least the first 2.5 months of the salmon’s marine life (80 km of the migration route).

Packing fish together in pens that are open to the sea is an almost ideal mechanism for breeding and distributing parasites and disease. In nature, you would never find salmon packed 25,000 to an acre. Keeping them in such conditions – and making them grow as quickly as possible – generally requires chemical manipulation. The earlier discussion here about antibiotic use and its role in the emergence of resistant bacteria is relevant.

These concerns also exist in addition to the fundamental reason for which fish farming cannot be sustainable: it relies on catching smaller and less tasty fish to feed to the tastier carnivorous fish that people enjoy. It thus lets us strip the sea bare of salmon or cod or trout and compensate for some period of time by using cheaper fish as a factor for their intensive production. Given that those cheaper fish are caught unsustainably, however, fish farming simply delays the emergence of truly empty oceans. And the industry is trying to have farmed salmon labelled ‘organic.’ Ludicrous.

Source: Krkosek, Martin et al. “Epizootics of wild fish induced by farm fish.” Proceedings of the National Association of Sciences. October 17, 2006, vol. 103, no. 42, 15506-15510.

P.S. Shifting Baselines also has some commentary on sea lice and salmon farming.