Home > Daynotes Home > Week of 25 February 2008

Photograph of Robert Bruce Thompson
Daynotes Journal

Week of 25 February 2008

Latest Update: Saturday, 1 March 2008 09:48 -0500

Paying for this Place
Visit Barbara's Journal Page

Monday, 25 February 2008
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

08:10 - Insurance Fears Lead Many to Shun DNA Tests. And they're right to be afraid. Insurance is, or is supposed to be, about spreading the economic consequences of risks.

The entire concept of insurance is based on the idea of a pool, the members of which are fungible insofar as the level of risk. For example, actuaries can state with high accuracy how many of a pool of 100,000 60-year-old men will die in the coming 12 months or how many of a pool of 30-year-old women will become pregnant in the coming 12 months or how many 17-year-old boys will be involved in car accidents.

The risk can be further quantified according to significant characteristics of the group in question. For example, actuaries know that more of the 60-year-old men will die among the subset of that pool who smoke cigarettes than among the non-smoking members of that pool, and they know with a high degree of certainty how many more. That allow them to set life insurance premiums that take into account the differing degrees of risk. Similarly, among the 30-year-old women, the actuaries know that a higher percentage of those who are married will become pregnant, which allows them to set health insurance premiums to cover those differing degrees of risk. And among the 17-year-old boys, they know that that subset of boys who have gotten speeding tickets are much more likely to be involved in an accident, which allows them to set car insurance premiums to cover those differing degrees of risk.

But what if there were no need to depend on the law of large numbers? What if, like Robert A. Heinlein's Doctor Pinero in Life-Line, there was a way to determine exactly which of those 60-year old men would die, or exactly which of those 30-year-old women would become pregnant, or exactly which of those 17-year-old boys would be involved in a car accident?

If that were true, the whole concept of insurance collapses. There's no point to distributing the risk if it's known ahead of time exactly which members of the pool will suffer a loss. Why should (or would) the vast majority of 60-year-old men who are not going to die in the coming 12 months pay higher life-insurance premium (or any premiums at all...), which would simply be a subsidy to the families of the unfortunate few 60-year-old men who are already known to be doomed in the next 12 months? Why should (or would) the vast majority of 30-year-old women who are not going to become pregnant in the next 12 months pay higher health insurance premiums, which amounts to a simple subsidy for the women who are going to become pregnant? And why would anyone pay higher car insurance premiums, or any premiums at all, if he knew he wasn't destined to be in a car accident?

These DNA tests are essentially such foreknowledge, and the question becomes how insurance companies and the government will deal with such foreknowledge. It's pointless to pretend that it's insurance. It's not, or at least not in the way insurance has historically been defined. Insurance companies are, or should be, entitled to use the best information available to set rates. Unfortunately, that means they'll refuse to insure people who are destined to suffer expensive illnesses, or at least write exclusions into the policies.

So, someone who has a DNA test that establishes that he's certain to suffer a particular type of cancer, say, will find that he cannot obtain health insurance that covers that condition. His neighbor, who is not destined to suffer cancer, could easily get cancer coverage although that would obviously be useless to him, but he won't be able to get coverage to cover the cardiac care that will be needed when he inevitably suffers a heart attack. In effect, we go from the historical situation where "we're all in this together" to an utterly different situation, where it's "every man for himself".

Now, DNA tests are not 100% certain, but 100% certainty is not required. If a DNA test establishes that you're ten times more likely than the general population to suffer from a particular condition, or even only twice as likely, you're not going to be able to buy "insurance" against that condition, at least not at a price anyone can afford. Why would any insurance company sell you such a policy, knowing ahead of time that they'll have to pay out on it? Or, if they did sell you such a policy, it would really just amount to a prepayment of the costs they know they'll incur.

Eventually, as such DNA tests become more refined, the whole concept of health insurance will go away. But that's only one of the impacts such tests will have. I can envision more profound effects. For example, in the interest of reducing future medical costs, the government may require that a couple be tested for genetic compatibility before they're permitted to marry or have children. People who carry bad genes may not be permitted to have children at all. For that matter, individuals may decide to make genetic testing for their prospective spouses a part of the prenuptial process.

All in all, that may or may not be a good thing. What concerns me is what won't be tested for. In avoiding pairings that pass down known bad genes, we may also be avoiding pairings that produce the next Leonardo da Vinci or Isaac Newton or Johann Sebastian Bach. Not because we're trying to avoid such combinations, but because we just don't know entirely what we're doing.

One thing is certain. DNA testing is here now, and will only continue to expand. Pandora's Box is open, and there's no way to turn back the clock.


Tuesday, 26 February 2008
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

09:28 - I'm still working on clean-up stuff for the home chemistry book. I wanted to include a periodic table of the elements, and it looks like we're going to do that as a full-color foldout on either the front or back cover. The periodic table will occupy one entire side of the two-page spread. The front or back cover will occupy half of the other side. That leaves me with a half side unused, so I have to come up with something useful to fill it.

I've also started ordering reference material, supplies, and equipment for the home forensics lab book. I already have quite a lot of the reference material, most of which is from the period 1890 to 1960 or so, when most forensics work was done using standard laboratory equipment. Nowadays, of course, a lot of forensics work is instrumental, but that doesn't mean the older "wet" techniques are no longer useful.

In fact, the majority of forensics tests involve wet chemistry even today, and even in wealthy countries. In poorer countries, where the expensive instruments used in modern forensics labs are unaffordable, forensics procedures haven't changed much from those used several decades ago. Even if you can't afford a mass spectrometer, for example, the Marsh Test is still as good as ever at detecting microgram levels of arsenic.

In fact, there are still situations where wet chemistry tests are preferred to instrumental analysis, particularly for toxins that are lethal in very small doses. Instrumentally, it can be very difficult to discriminate such toxins at the very low levels typically present in body fluids because the signal from the toxin is lost in the noise of everything else present. But forensic chemists in the last half of the 19th century and the first half of the 20th worked very hard to invent tests for many such toxins. These tests are in many cases extremely specific to a particular toxin and can detect it at very low levels. We'll be exploring some of that in the forensics book, although of course not with real toxins. Well, not many real toxins, anyway.


Wednesday, 27 February 2008
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

08:01 - I read a strange article in the newspaper this morning. It was talking about the increasing use of ethanol for fuel and the implications for fire fighters. According to the article, standard firefighting foam is useless for ethanol fires because the ethanol just eats through the foam and keeps burning. Special foam is needed, which costs a lot more than standard foam. Okay, I can believe that.

What I don't understand is the comment that water is useless for fighting ethanol fires. According to the article, burning 80% to 95% ethanol can't be extinguished with water, which seems very strange. Ethanol is miscible with water, and ethanol/water mixtures with less than 40% to 50% ethanol (depending on conditions) don't burn. So why shouldn't plain water extinguish an ethanol fire? I may have to test this to satisfy my curiosity.


Thursday, 28 February 2008
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

08:48 - The end of this week is also the end of this month. I'm not where I'd hoped to be, but at least I'm making good progress on all fronts. Much remains to be done to get ready for the home chem lab book arriving in the stores, but all of that is well in progress. The forums are set up and configured, although not yet active, and I've at least gotten a start on the support web site. I just sent the first draft of the teacher's guide to Mary Chervenak and Paul Jones to get their comments on it. Mary is coming over the weekend of 7 March to shoot the first in the series of homechemlab.com videos.

My editor at O'Reilly sent me the contract for the home forensics lab book yesterday, which I looked over, printed out, signed, and returned. I'm still in the early stages of that book, accumulating the equipment and reference sources I'll need. As usual, everything is a disorganized mess in terms of how the new book will come together. With my early books, that always worried me, but I'm used to it now. It's kind of like a sculptor starting with a block of stone. In the early stages, it doesn't look like much of anything, but eventually it all comes together.

The milestone dates in the contract are quite generous. For example, the first-two-chapters milestone is something like 15 May. I intend to beat those dates, and by quite a lot. I intend to finish two more DIY Science books this year, the home forensics lab book and one other that's not yet determined.

11:00 - I see that now some people are trying to claim that McCain is not eligible to serve as president because he was born in Panama. I don't particularly like McCain, but that's simply outrageous. McCain was born a natural US citizen, of parents who were both US citizens. Here is the relevant text:

"No person, except a natural born citizen, or a citizen of the United States at the time of the adoption of this Constitution, shall be eligible to the office of President;"

Which seems definitive to me. You'll note that the text doesn't say that one must be born within the borders of the United States, but merely that one must be a "natural born citizen". McCain qualifies on that basis. He was born a US citizen, not naturalized.


Friday, 29 February 2008
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

08:58 - Leap Day falls on February 29th this week.

Wednesday, I invited Mary Chervenak to join us in our traditional Leap Day celebration, dancing naked around our large pine tree. She said she'd join us if the weather forecast was favorable. She didn't define "favorable". Barbara says she plans to stay indoors this evening. Women are such wimps.

I'm working now on one of the labs from the home forensics lab book.

Laboratory 12.1: Analyze Ethanol Content in Exhaled Air

Which is basically a reproduction of the original 1954 Breathalyzer test. The redox reaction between ethanol and potassium dichromate in acidic solution converts the yellow-orange chromium(VI) ion to the blue-green chromium(III) ion. The solution starts out yellow-orange. When air containing ethanol passes through the solution, some or all of the yellow-orange chromium(VI) ions react to form blue-green chromium(III) ions. The greener the solution becomes, the more ethanol was in the sample. The original Breathalyzer machine used photometry to detect the amount of change. My readers won't have access to a photometer, so we'll use visual photometry, with several samples of known ethanol concentrations for comparison.

Breathalyzer machines, original and current, use breath alcohol content as a proxy for blood alcohol content, which in turn serves as a proxy for degree of impairment. There are two obvious problems with those assumptions.

First, breath alcohol content does not necessarily correspond to actual blood alcohol content. You can, for example, reduce the value displayed by the Breathalyzer by as much as 50% simply by hyperventilating for 30 seconds or so before the test. Alternatively, by holding your breath for 30 seconds, you can produce a false reading that's 50% or more higher than actual.

Second, even actual blood alcohol content correlates only very loosely with the actual degree of impairment, which differs dramatically between individuals. I remember 10 or 15 years ago when MADD or some such organization decided to demonstrate the actual degree of impairment with increasing BAC. They made an event of it, with local TV news cameras present. They set up a slalom course with traffic cones and got NASCAR racer Kyle Petty to volunteer as the driver. Over the course of several hours, they had Petty drinking shots of bourbon, doing Breathalyzer tests after each shot, and then having Petty try to drive as fast as possible through the course without knocking over any traffic cones.

The problem was, their demonstration showed exactly the opposite of what they intended. The drunker Petty got, the faster he made it through the course, and the fewer traffic cones he knocked over. By the time Petty was well over the legal limit, he was making it through the course in record time without knocking over any cones. If they'd used me as their test subject, the results would have been different. After one beer, I'd have been knocking over cones, and after two beers I'd probably have just plowed right through all of them. But then I typically drink only a few beers a year, and I have no tolerance built up for alcohol.


Saturday, 1 March 2008
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

09:48 - I've mentioned my standard backup procedure before, but as I was doing it this morning I thought some might be interested in the details. My Linux guru friends no doubt think my method is incredibly convoluted and clumsy. More than one has pointed out that it could easily be automated, but I'm more comfortable doing it myself manually every morning. It takes only 10 minutes, and that's 10 minutes I'm willing to spend to make sure my backups are available and readable when I need them. Here in detail is what I do:
  1. Take one of my Kingston USB sticks back to Barbara's office. (I could obviously back up Barbara's system over the network, but doing it this way forces me to sit down at her machine once a day, where I can check for updates and so on.)
  2. Fire up Konqueror on Barbara's machine, which displays her home directory.
  3. Select the top item (.adobe) and press Ctrl-A to select the contents of the entire directory. Right-click on one of the highlighted items and choose compress to a tar.gz archive.
  4. Plug in the USB stick, which automatically pops up another Konqueror window. Cut archive.tar.gz from Barbara's home directory and paste it to the USB stick.
  5. Pop a console window and press up-arrow to insert the most recent command, which is always "sudo sync". Press Enter and wait until the data is synced to the USB stick. Repeat "sudo sync" just to make sure.
  6. Close both Konqueror windows and the console window.
  7. Right click on the Kingston USB icon and choose Remove Safely (yes, I know the sync command did that, but why not be extra safe?)
  8. Remove USB stick and carry it to my office.
All of that takes about two minutes, literally. I then head for my office, where I:
  1. Double-click the Working icon on my desktop to fire up an instance of Konqueror that displays /home/thompson/data/working
  2. Double-click on /home/thompson/data/working/backup-barbara to display that directory.
  3. Insert the USB stick, which automatically fires up another instance of Konqueror. Copy/paste archive.tar.gz from the USB stick to the backup-barbara directory.
  4. Fire up a console window and up-arrow to run "sudo sync" twice.
  5. Close the USB stick Konqueror window, right-click on the KIngston USB icon, choose "Remove Safely", and remove USB stick.
  6. Double-click the Thompson icon on my desktop to fire up an instance of Konqueror that displays /home/thompson
  7. Highlight the top item in that directory, and press Ctrl-A to select all items in that directory.
  8. Scroll down and use Ctrl-click to deselect the directories .wine, data, and junk.
  9. Right click on one of the selected items and choose compress to a tar.gz archive.
  10. Change the Konqueror instance from /home/thompson/data/working/backup-barbara to /home/thompson/data/working/backup-robert
  11. Cut archive.tar.gz from /home/thompson and paste it to /home/thompson/data/working
  12. In the console window, up-arrow to run "sudo sync" twice, and close Konqueror window with /home/thompson
  13. Double-click the Backup icon on my desktop to fire up an instance of Konqueror that displays /backup (which is on a second physical hard drive).
  14. Change into the usrback directory, which contains one subdirectory for each day. Create a new directory in the format /backup/usrback/20080301 Saturday. Change into that directory.
  15. Copy the /home/thompson/data/working directory and paste it to the daily directory on /backup
  16. In the console window, up-arrow to run "sudo sync" twice.
  17. Change up one directory and select the daily directory
  18. Fire up one of my external hard drives. When it mounts, it automatically fires up a Konqueror instance that displays its contents.
  19. Change into the usrback directory.
  20. Copy /backup/usrback/<today's-directory> and paste it to usrback on the external drive
  21. In the console window, up-arrow to run "sudo sync" twice.
  22. Close the Konqueror window for the external hard drive.
  23. Highlight the icon for that drive and choose "Remove Safely"
  24. Power down the external hard drive.
  25. Repeat steps 18 through 24 for the second external hard drive.
I do that every day, Monday through Sunday. On Sundays, I also burn a copy of /home/thompson/data/working to a DVD+R disc, as well as whatever happens to be awaiting archiving in /home/thompson/data/holding

If that all sounds incredibly complicated and time-consuming, it's not. I can do it almost without having to think about it, and it takes 10 minutes or less per day. Again, that's 10 minutes I'm willing to spend to ensure the safety of my data. The two external hard drives are in external enclosures with the covers off. If I'm leaving the house for anything longer than walking the dogs to the corner and back, at least one and usually both of those hard drives go with me. It takes about five seconds to pull one of the hard drives from the enclosure, stick it in one of the clear plastic cases that Seagate uses to package hard drives, and put it into my Lands' End attache.

I also periodically dupe the entire primary hard drive of my main desktop system to another hard drive, which also goes along with me. If worse came to horrible, all I need is any PC with a couple of SATA connectors, and I can be back up and running in the same state I was that morning in literally five minutes.

Oh, and during the work day I routinely copy whatever document(s) I happen to be working on up to the server that runs this site, so the most I can ever lose is perhaps one hour of work.


Sunday, 2 March 2008
[Daynotes Forums]    [Last Week]   [Mon]  [Tue]  [Wed]  [Thu]  [Fri]  [Sat]  [Sun]   [Next Week]    [HardwareGuys Forums]

00:00 -


Copyright 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008 by Robert Bruce Thompson. All Rights Reserved.