Friday, May 30, 2008

IT skill shortage

I keep trying to write a full post on this, but get discouraged by the magnitude of the task.. As a placeholder, here's a short response, provoked by this article.

If there were a skills shortage, then IT salaries would be increasing, not decreasing.

There is also no actual data to support the contention that there is a shortage of trained people. See the Business Week review of the data.

So what the executives call a 'skills shortage' must mean something different. I suspect this means 'short-term shortages of engineers with specific technical skills in certain industry segments or in various parts of the country', as the second article above notes. That is, the executives find it inconvenient that they cannot snap out a C# engineer and snap in a Cobol one, or vice versa, as their needs dictate. My sympathy is limited.

The answer is really as simple as Sandra outlines in the first post - invest in technical staff, to get them the training they need to acquire the skills. The mainframe is not a mythical beast, with its habitat and behaviours shrouded in mystery: it's well-documented and easily learnt. It's the attitude to technical staff that classifies them as interchangeable parts, which produces the appearance of the 'skills shortage'.

Of course required reading on this is Dr. Matloff's investigations.


The redoubtable Yakov points us, grumpily, to Jason on 'what to do when you're laid off'. Jason thinks the appropriate response is to book a flight to somewhere cheap for a long vacation. Yakov does not.
I am of Yakov's mind - this advice is good only for the glittering digerati, the young strong and lucky, the children of privilege. I can't speak for Yakov, but this might be an emigrant thing. Those of us who escaped only by dint of outworking the competition, have a jaundiced view. We see a steep and thorny way through the miseries of poverty: salvation is through hard continous work.

Of course it may also be simpler, maybe it's just the optimist/pessimist divide. I have a nagging suspicion that those damn'd cheery optimists have a way of creating good luck by their very sunniness. I've never been able to fake optimism though and certainly can't pretend to the real thing, so this remains a source of gloominess. Hah.

One more thing: quoth Jason,
"But I'm married with a family and a house... Ok, you win. You're screwed, but that's the life you chose for yourself so you're going to have to live it. It's worth noting, however, that most Europeans wouldn't consider that a reason not to travel. Right this second, there is a German couple pushing a stroller down a remote beach in Thailand, and they're not going home for another month. What's your excuse again? "

Well, let me count the differences between a US couple and that German couple. The German couple each have a Worker's Council (read: union) to protect their interests at work; they have 5 to 6 weeks of paid leave a year, plus 10 other paid days off; they have state-guaranteed health care, so if they lose their jobs they don't lose health care coverage; they have state-guaranteed pensions, so they don't have a 401k as the only thing between them and dogfood for dinner in retirement; they have the euro, the strongest currency on the planet: five excellent reasons they can be a tad more carefree than US wage slaves. "Americans average 25.1 working hours per person in working age per week, but the Germans average 18.6 hours."

That German couple is on vacation for a month. The only way a US couple could get a month's vacation, is to be laid off simultaneously: at which point their family is laid bare to the rapacious wickedness of ill fortune.

Wednesday, May 28, 2008

bold Nebraska

We were in Nebraska for Memorial Day. Drove 6 hours on Friday to cover 280 miles, with 30-40mph crosswinds tugging at the canoe the whole way. Pulled into North Platte in horizontal rain, Jeff saw tornados coming down from the clouds. We sat in a restaurant and discussed options. Hotelling it was the consensus, got a 2-bedroom hotel suite, the wives and children were very happy. After the kids had swum for an hour in the hotel pool, we watched kayaking on TV, how pathetic. 5 inches of rain and 60mph winds in our destination town of Valentine that night, so I think it was the right decision.

Next day attempting to pitch the monster 8-man tent in a Nebraska light breeze, started setup with tent backed to the wind: by the time it was up the wind had swung through 180 degrees, then a strong gust blew up the tent. I mean literally blew up, the flysheet exploded along the seams, pole shattered, splitrings holding pole anchor points on the inner were straightened out.. etcetera. Then it blew away, looked like a giant tumbleweed thundering through the campsite, nearly took out some tourons walking through to see Smith Falls. Once we'd subdued it again, drove into Valentine for a new tent. Young's Mercantile didn't sell tents but recommended Fred's Bait Shop, Fred had only 2-man tents unfortunately. The hardware store had a 5-man luckily, a mere $160. Next time I'm just checking straight into the hotel and staying there, it will be cheaper.

River high and chilly. On Mem. Day it was a dark and stormy morn, ranger came up as we were packing and said 'you can't leave now, it's just about to start raining again'. Everyone's a humorist.

bold Nebraska ? the song is always in my head when driving to NE. But I see the lyric is actually 'all aboard, Nebraska's our next stop' where I'd always heard 'bold Nebraska's our next stop'. I like my version better. This turns out to be a mondegreen

Wednesday, May 14, 2008

vaccines and autism

CBS interviewed Dr. Healy, and now seems to think this is an 'Open Question'. From the interview,
"why in the past decade hasn't the government compared the autism/ADD rate of unvaccinated children with that of vaccinated children?"

Because this has been done before, many times. No link between vaccination and autism has yet been found.
From an article in the Guardian:
"The Danish Epidemiology Science Centre compared 440,000 children who had MMR with 97,000 children who didn't. The children who had MMR were no more likely to develop autism than the children who didn't. A group in London looked at 498 children with autism, to see if they developed it after MMR. They looked at when they had the MMR jab, and when they developed the symptoms or the diagnosis, and found no sudden blip after immunisation. Another paper shows no increase in GP consultations in the six months after immunisation. Two hundred children in London and Stafford with autism were studied to see if there was a new type of autism related to MMR, featuring bowel problems and sudden regression, a bit like in the drama: half had the jab, half didn't, and there was no difference in type of autism between the groups. In California, looking at 1,000 children a year, over 14 years, the number of cases of autism increased by 373%, while the number of children getting MMR increased by only 14% (from 72% to 82%). "

If anyone has published studies showing evidence of a link, I'd surely like to see it.

From the interview again,
"If we can screen children to see which ones might be more susceptible to vaccine side effects.."
There is no known theory and no plausible biological mechanism for vaccinations to cause autism. So how could this screening be done ?
At one point thimerosal was postulated as a possible link. That's a separate discussion, but irrelevant now: child vaccines containing thimerosal are no longer used. Since then, there hasn't been any new causation mechanism proposed.

Also, a study in Japan, published in New Scientist print edition, 17 February 2001:
Journal reference: Journal of Child Psychology and Psychiatry (DOI: 10.1111.j.1469-7610.2005.01425.x)
"They found that the number of children with autism continued to rise after the MMR vaccine was replaced with single-shot vaccines. The medical records of 31,426 children in the city of Yokohama were checked. Before the vaccine was withdrawn, between 48 and 86 children per 10,000 were diagnosed as autistic. After the vaccine was withdrawn, 97 to 161 children per 10,000 were diagnosed with the condition."

There is considerable evidence that the "increase" in autism rates is an artifact of better diagnosis. See the Journal of Pediatrics, and the British Medical Journal.

If you want scary, contemplate thousands of children in iron lungs with polio, thousands of children blind or brain-damaged by measles, epidemics of whooping cough and thousands dying, malformed babies due to their mothers being exposed to rubella, etcetera. This is a certain result of large-scale refusal of vaccinations for populations at risk.

While Dr. Healy has undoubtedly done a lot of good work in the world, I suspect there may be a political aspect to this. Note that Dr. Healy is a Republican political appointee, and was "a member of the Advisory board of The Advancement of Sound Science Coalition, an organization later shown to have been founded by PR firm APCO and funded by the Phillip Morris corporation to criticise scientific research inimical to the interests of tobacco companies and other corporations" (Wikipedia). I don't see what advantage there is to starting this particular hare again, but I suspect it's there. It may just be part of the broader front in the Republican war on science.

The benefits of vaccination are proven. The hypothesis of a link between vaccination and autism is speculative and has no theoretical or evidentiary support. That doesn't mean it does not exist, only that it's unlikely in the extreme. From a public health perspective, investing in studies of a speculative unsubstantiated hypothesis is not easily justified.

Update September 2008: hm, now it appears Lance Armstrong is going anti-vaccine. Science saved his life and career, but now he's joining the druids, the forces of old night and chaos.
Further reading on the subject, rather better-informed than my brief jog through it, here and elsewhere on the site. Predictably, measles is first to make a comeback, but the other horsemen aren't far behind.
Some measurements of the number of lives saved by vaccines is at

More on the horrorshow that preceded immunizations at Making Light.

Thursday, May 8, 2008

asthma and doping

Petacchi, the dominant sprinter of the last few years, has been suspended until August by the not-august Court of Arbitration for Sport: for a doping violation involving the asthma drug salbutamol (yclept albuterol in the USA).

In the article, it's claimed that 'Salbutamol can be used to increase an athlete's anaerobic power. '

I have adult-onset asthma, and take salbutamol when racing. The reason I noticed the asthma was that my 5k run times were 10-20% slower at maximal effort, than what training results predicted, with the addition of unusual distress. My training logs go back to 1975, so there's enough data there that I at least am satisfied that the predictions are reasonable. Tests confirmed the diagnosis, so I take it as prescribed now. With salbutamol, performances are back to normal: but I have never been able to measure any performance enhancement, even when running short maximal-effort intervals (400m repeats).

So I took a closer look at the data. The studies that show improvement in anaerobic power used a dosage of 12 mg/day for four weeks. See

That's 12 milligrams a day. To put this in perspective, an inhaler delivers about 200 micrograms per dosage. The anaerobic power improvements require 12 000 micrograms per day. This is equivalent to 60 metered doses from an inhaler. That's a dose every 16 minutes (assuming 8 hours sleeping), for four weeks. It's impossible to hit this dosage except by taking it orally or intravenously. It's clear that Petacchi wasn't doing this.

There are also lots of studies showing that at dosages that can be delivered via inhaler, there are no measurable performance improvements, aerobic, anaerobic or anabolic.
See for example
- The effects of albuterol on power output in non-asthmatic athletes, LEMMER J. T et al,
International journal of sports medicine ISSN 0172-4622 1995, vol. 16, no4, pp. 243-249 (33 ref.)
- The effect of salbutamol on performance in endurance cyclists, Norris Petersen and Jones,
European Journal of Applied Physiology ISSN 1439-6319 (Print) 1439-6327 (Online), Volume 73, Numbers 3-4 / May, 1996

Then there is the question of how the salbutamol dosage is measured. WADA's measure is 1000 nanograms per milliliter (ng/ml) in the urine. This is not a very good measure. See for example this study, which notes:
"Urine cSAL increased with dose and was highly variable, with the peak value observed being 831 ng x mL(-1) after a dose of 800 microg.
Conclusions: Inhaled SAL does not enhance time-trial performance, regardless of dose, and that urine cSAL after exercise is related to dose, demonstrates high variability, and is partially related to hydration status."

Even at 800 micrograms, an order of magnitude lower than a single day's dose of the amount required to produce the anaerobic improvements (and that dosage was repeated for four weeks), it's possible to get very close to the WADA limit. Add in variability, plus the fact that dehydration will artificially elevate the measurement in the urine, and it's clear it's possible to exceed WADA's limit without seeing any performance improvements.

There are a couple more problems here. Metabolization will differ by individual, and by the environment in which it's taken - a fat breathless desk jockey like me is likely to have a different rate from a Tour de France sprinter. Metabolism of salbutamol differs between asthmatics and healthy people too, with the asthmatics showing higher concentrations of salbutamol in urine for the same dosage.

Another thing that's missing is data on the urine concentrations in the subjects who took the massive doses, and showed anaerobic performance improvements. Do they show the same variability of urinalysis data ? What kind of urine measurements were found ?
I suspect their urine would show levels massively higher than the 1000 ng/ml that is currently considered a violation.

Personally I give any endurance athlete the benefit of the doubt in a salbutamol case. A cycling sprinter like Petacchi is a special case, since he could theoretically benefit from the anaerobic performance improvements: but he still has to ride 4-6 hours at moderate to high aerobic pace, just to reach the sprint. Anaerobic performance after this kind of effort has never been tested. Add that to the questions around the current level which is considered a violation, and I think all these salbutamol 'violations' are likely bogus. The potential benefit is too small for the risks.

This 'violation' in particular is nonsense. Even the CAS said, "the adverse analytical finding in this case is the result of Mr. Petacchi simply, and possibly accidentally, taking too much Salbutamol on the day of the test, but that the overdose was not taken with the intention of enhancing his performance. Indeed, it would be an unusual way of attempting to enhance performance to take the prohibited substance after the particular event had concluded."
So.. what is this ? zero-tolerance, as propounded by the imbeciles that propagate the War on Drugs ? It's senseless.