I voted in a municipal election in Toronto, Canada earlier this week. Not on the voter's list? No problem--you can register at one of the city clerk's offices. There's five of them, serving a population of 2.6 million people. Oh, and they're open from 8:30am to 4:30pm, Monday to Friday. So that should be a snap to get to, as long as you don't have a full-time job, or a child to care for, or mobility issues. (You don't mind choosing between a couple of extra bus fares and eating lunch, do you?)

I followed the campaign closely, I was aware of the major issues of the day (as well as the minor issues that didn't get nearly enough coverage), I had strongly-held opinions based in thorough, extended research--and I registered to vote on the day of the election.

The notion that all people who didn't register in advance are somehow lazy, unworthy, and incompetent is canard that punishes the working poor, the single parents, the handicapped. Looking in from the outside, it's apparent that it's one piece of a larger Republican campaign to disenfranchise as many Democratic-leaning voters as possible. It's a story that is propagated by Fox News, the viewers of which are exemplars of the Dunning-Kruger effect.

You know, I just put together now that "SJW" is intended to be an acronym for "Social Justice Warrior" (which is in turn intended to be a derogatory phrase meaning, as far as I can tell, "uppity feminist"). For some weeks now, I have been pondering what the internet has against straight (or single) Jewish women. Now it makes a lot more sense.

That the "reasonable" faction of the male IT world - that the parent poster would like to think he represents - seems to believe that the SJW caricature represents a non-trivial force that is conspiring against him is troubling. That the acronym SJW exists and is presumably widely understood in his circles is rather more revealing about (his part of) "IT culture" than he probably thinks.

Don't get me wrong, the parent poster is better than the grandparent asshole who believes all rapes are imaginary--but just being better than the anonymous trolling asshole isn't setting a high bar.

When it's an infraction that is easy to avoid? Yeah, sign me up. (And what's this "permanent unemployment" nonsense?)

No one accidentally has a beer, and no one accidentally gets behind the wheel of a car. If there were a way to ensure that selfish assholes only put their own lives at risk, that would be one thing--but this situation isn't that. Incidentally, I feel the same way about the people who think they're still good drivers when they're on their cell phones. (To be clear, that's everyone who is driving while using a portable electronic device. No, you aren't special.)

Not quite true. You have to be white, and preferably wealthy or a member of a police force, and preferably directing that violence toward a person of color. Being from a red state helps, too.

To give the devil his due, the best locations for solar installs tend to be sites that aren't very valuable for agriculture, residential, commercial, or manufacturing use. They're out in the desert. This isn't to say that desert land is valueless (economically or environmentally) but generally it is a type of land that - until now - we have had very little incentive to exploit, and there is an awful lot of it.

There's a limit to how good things can get on the solar panel side. The best multi-junction photovoltaic cells, at the cost of great complexity, are able to reach about 45% efficiency in the laboratory. The absolute maximum theoretical efficiency is about 85% (requiring materials and manufacturing processes that haven't been discovered and probably don't exist). One server per square meter is still pitifully low density. On the bright side, the cooling problems get to be much easier to deal with, I guess.

As for the per-server power draw decreasing--that's true, though it's happening slowly. And the trend has certainly been to increase the number of servers in a data center (or the number of blades in a rack) much, much faster than the per-server energy consumption has gone down.

This isn't to say that I'm opposed to rooftop solar. The balance between rooftop area and demand is much more favorable if you look at, for instance, suburban homes (which are admittedly otherwise environmentally disastrous). And rooftops are 'dead' space otherwise, that might as well be doing something useful like producing electricity. My point was only that the suggestion that all would be solved if the data centers put solar panels on their rooftops (or even their parking lots and windows) was nonsense. And, incidentally, that the most efficient places to put solar generating facilities are actually a long way from where the majority of people are living.

Above the atmosphere, at the equator, the average insolation (that is, the amount of incoming solar energy, averaged over the course of a day) is about 400 watts per square meter. At the bottom of the atmosphere in an ideal location (like the Sahara) it's closer to 300 W/sq. m. In most places where people want to have data centers, the number is closer to 200 W/sq. m...or worse. And the efficiency of commercial solar panels runs about 20%, so you're down to 40 watts per square meter.

200 watts is (optimistically) about the draw of a single server, so you're looking at powering one server for every five square meters of rooftop. If you want to run on rooftop solar, then you're going to have to design a data center with very short racks and very wide aisles.

Indeed. One wonders what remarkable scientific discoveries and conclusions will result from creative analysis of today's data, forty years hence.

...At least, as long as we actually do continue to fund new instruments and research, and don't insist that all data collection is now 'done', and that all the work that remains is winnowing smaller and smaller pieces of useful information from the last century's scientific output.

No ,they don't need a minimum level of proof to carry out the seizure. They need a minimum level of proof to defend the seizure in court--which is a totally different ball game. Attorneys cost money, even if fees are eventually awarded many potential plaintiffs can't afford to be out of pocket for the time (months or years) required for a case to make its way through the courts. Seizures made against out-of-town and out-of-state victims are even harder to challenge--it can be quite costly to repeatedly travel to a distant jurisdiction's courts, even if you can afford to take the time off work. And to challenge even a blatantly illegal seizure is to invite additional scrutiny and future harrassment.

If crooked cops can hit the 'sweet spot' of around a few thousand dollars, in most cases it's going to be too much of a hassle and expense for a victim to fight.

The reason is because the Icelandic alphabet has two different letters that produce a sound that could be written "th" in English. The letter eth (Ð or ð) is a voiced "th", like in "they" or "this"; whereas the letter thorn (Wikipedia link, since Slashdot won't render it) is unvoiced, like in "thistle" or "theater". By convention, eth is transliterated as "d", whereas thorn is transliterated as "th". It does make some sense, as "d" is a voiced consonant, so that in addition to the look being similar, the naive pronunciation isn't horribly wrong. (And it means that someone seeing a "th" in an English transliteration of Icelandic text knows that it's unvoiced, so they'll get the pronunciation right.) And for better or for worse, it's the accepted transliteration, so if you want to fight it you're fighting against convention.

The gross transliteration error that kills me is when someone substitutes P or p for the thorn and turns something like Thingvellir into Pingvellir. That's just horribly wrong.

Exactly correct. If the target date for an "interim test storage site" is 2021, that's only 7 years out.

Let's allow a year to figure out what the specs ought to be, a year to request and evaluate proposals from possible contractors, a year to build prototypes, a year of testing, a year to fix problems identified in testing, a year to manufacture the first few final-version railcars, and a year for overruns. That's a seven-year timetable right there.

Unless we want to be running late, paying tons of money out in overtime, and getting railcars that kind-of-sort-of work right most of the time...then yeah, right now is a good time to start on this stuff.

If the $4.4 billion price tag for the San Onofre facility is anywhere near the right ballpark, a $30 million expenditure 35 years down the road would be, in today's money, a rounding error.

It also should go without saying that we do have 30+ more years of experience with decommissioning nuclear facilities now than we did in 1979. And San Onofre, unlike TMI, was not the site of a significant accident that damaged its core and contaminated the facility.

Wages - and especially that subset of wages which are paid at the legal minimum - represent only a fraction of the total costs of operating a McDonald's restaurant. All wages together are about 25% of the total costs, and that includes a non-trivial number above-minimum management and support staff. So even if we make the unreasonable worst-case assumptions that a) all employees do earn minimum wage, and b) that increased wages don't result in any improvement in average employee productivity (because employees are physically healthier and because of reduced turnover) then a 1% increase in minimum wage only makes for a 0.25% increase in cost-of-Big-Mac.

And a 0.25% increase in cost-of-Big-Mac doesn't actually equate to a 0.25% increase in actual cost-of-living. The effect will be smaller or negligible for businesses where staff costs represent a smaller share of total costs, and where dealing with businesses in which employees are already better paid than minimum wage.

And finally, there are a number of costs associated with minimum-wage workers that you're already paying out of your own pocket, without realizing it. Wal-Mart and McDonald's know perfectly well that minimum wage isn't a living wage. Food stamps, state-subsidized health insurance programs, school lunch programs--that's money you're paying because Wal-Mart isn't. Forcing McDonald's to pay its employees a living wage (or closer to one, at least) means that your Big Mac's price is (less) subsidized by the government.

Used to be more true, not so much today. The Highway Trust Fund - which is funded by a combination of federal fuel and vehicle taxes - has been bailed out before ($35 billion between 2008 and 2010) and is out of money again this year. And the federal government has turned over responsibility for the interstate highways to the individual states, so a big chunk of the construction, maintenance, and repair bills actually comes from the states.

Looking at 2010 numbers, total spending nationwide on highways was about $155 billion. The federal gas tax brought in $28 billion; state and local fuel taxes amounted to another $37 billion; plus state and local governments picked up another $12 billion from tolls and non-fuel taxes. All in all, that's about $77 billion in revenue for $155 billion in expenditures. Drivers are paying about...51% of the cost of the highway network.

For comparison, I note a comment below that shows in fiscal 2012 Amtrak spent $4.036 billion and had revenues of $2.877 billion. In other words, Amtrak riders paid 71% of their costs out of pocket--a much bigger share of the costs than highway users.

Actually, not so surprising, depending on how the analysis is done. And it also depends a lot on how you want to measure "sheer productivity". A supervisor who helps design the experiment, interpret the data, write the paper, and communicate with journal editors probably spends fewer hours than the trainee (grad student or postdoc) who actually does all the bench work--but that doesn't mean that the supervisor hasn't earned an authorship credit.

If Alice, Bob, Carol, Dave, and Elsa are all graduate students in Dr. Frink's lab, and each of those students publishes two papers over the course of their PhD programs, then all of those students are going to be authors on 2 papers each, and Frink will be an author on 10 papers. Dr. Frink is 1 out of 6 scientists - a bit less than 17% - but is on 100% of the papers. If you have a big lab in a relatively hot (or well-funded) field, then your name is going to be on a lot of papers.

And papers these days - especially the high-impact, widely-read, highly-cited papers - tend to have a longer list of authors. If you look at the table of contents for the most recent issue of Science, the two Research Articles have 26 and 12 authors. Out of the dozen or so Reports, one has 4 authors, two have 5, all the rest have more. Speaking personally and anecdotally, my last three manuscripts (in the biomedical sciences) had 8, 3, and 7 authors.

Going back to "1% of scientists are on 45% of papers"--well, if those are all six-author papers, then that top 1% is only responsible for a 7.5% share (45 divided by 6) of the "output". Given that there is a very long tail of authors who only have 1, 2, or 3 authorships in their lifetime (the majority of PhD graduates never end up conducting research as university faculty; there just aren't enough jobs), I am willing to believe that there is a small fraction of productive, top scientists whose names are on a disproportionately large share of papers.