I'm losing badly in my NCAA pool. This may partly reflect my methods: I followed the seeding blindly, and then randomly added upsets until my "Madness Level" reached the historical norm. I was solidly in the middle for the first round or so, and since then have sunk steadily to my current position second to last. Gus started near the bottom but has slowly worked his way up, and key to this has been Missouri: he picked his home state to go one round further than pure seeding suggests, and then abandoned them in favor of top seed UConn in the fourth round. He called this an "emotional hedge". I send the following email to the pool participants this morning, and felt very satisfied that, before 11 am on a bright, not-cold, not-windy spring morning, I'd already ridden 8 miles and written the best email of the day.

Gus contemplates the moral vacuity and, we might even say, spiritual nihilism of placing an "emotional hedge" against his own soil and heritage. Kona observes, non-judgmental in all matters not involving food.

My satisfaction was premature, however, because moments later I sent out the new best email of the day:

Gus contemplates himself contemplating his moral vacuity. Searching for his tattered, damned Catholic soul, he finds only the University of Connecticut. With no food at hand, Kona gives Gus his space.

The Gamble

By Washington Post reporter Thomas Ricks. Ricks' reporting posits that, first, US military activity in Iraq in 2003-2006 was so utterly counterproductive to US aims that Bin Laden couldn't have asked for anything more. He covered that in his previous book, so this book is about "the surge". The surge is several things first, sending all available US troops to Iraq (amount to roughly 35,000, but the surge architects would have takes double that if the troops had been available). Second, a major change in US tactics, directed by General Petreus and amounting to using counter-insurgency methods , as opposed to ... not using counter-insurgency methods. The key elements of the new tactics are bribing Sunni insurgents to change sides, and changing priorities from trying to kill "bad guys" to trying to protect the Iraqi civilian population. The surge also benefited from Moqtada al-Sadr's decision to reduce his militia's level of attacks, and from the simple fact that most ethnic cleansing was already finished. The effect of the surge was to drastically reduce daily violence, but it leaves all of the major factions in Iraq completely unreconciled and fully armed, and may have simply postponed an inevitable civil war.

The book is extremely readable and I highly recommend it. It contains many details worth knowing, such as a thorough indictment of essentially every major military decision-maker from the Secretary of Defense to the Chiefs of Staff to the generals, from 2003 to 2006. Its main flaw is that it's simply the story of the second batch of generals, as told to this reporter. While those generals certainly seem to bring enough realism and pessimism to the story to make it plausible, Ricks is working through a very narrow perspective. He does mention negatives about the heroes of his story (Petreus and Odierno, primarily), and he is clear that he believes that while surge succeeded in buying time, that time won't be used for anything positive.

But the problem is not one of having enough bad things to "balance" the good things. As readers we depend on Ricks' judgment of what to mention in 300 pages to give an accurate impression of his years of research, both factually and emotionally. But the vast majority of his sources were either second-wave military leaders in Iraq and pro-war think-tanks, so he simply couldn't have enough information to make those judgments, no matter how plausibly realistic and self-critical the current military leaders seem to be, and we in turn must read between the lines for clues on how to weight Ricks' opinions. Ricks is as generous to his sources as he is harsh toward their predecessors, and his sources include some neo-cons, such as Fred Kagan, who clearly merit a far more critical approach.

Even with those caveats, the perspective he does provide is certainly a valuable one and one that belongs in any evaluation of the war, and the book is excellently done.

The current modules on the International Space Station are called:

• Zarya (sunrise)
• Zvezda (star)
• Unity
• Destiny
• Columbus Research Laboratory
• Kibo (hope)
• Quest
• Pirs (pier)

The next module going up, "Node 3", is the subject of a NASA naming contest, in which the write-in "Colbert" just beat the most popular NASA option, "Serenity". Serenity itself was popular in large part because it's the name of the spaceship from the tv show "Firefly" and movie "Serenity". NASA now has to either call the module Colbert, screwing up the naming convention, or override their poll. I propose a different solution:

Call it "Colberity".

Richard Stallman and the Free Software Foundation make a very good point, that if you browse to something like Google Office, it downloads a proprietary program onto your computer, so that even though you may be using only free software on your computer (e.g., linux, firefox) and the program is in an open language (javascript), you still end up with a non-free program. They go on to propose some simple conventions to help avoid this.

But in the middle of making the point, some inflammatory language appears:

The term "web application" was designed to disregard the fundamental distinction between software delivered to users and software running on the server.

I use the term web application myself, and I don't use it in order to disregard the fundamental distinction between local and remote software. I use it because it accurately describes most of the programs I work on, programs which are complicated enough that "web site" is misleading, and which users interact with via web browsers instead of local desktop applications. I don't feel like a useful idiot or dupe of nefarious forces who "designed" the term to obscure an innate distinction with implications for freedom. I suspect the people who coined and popularized the term had the same motive that I do using it: it is an accurate and terse description of a type of software. It wasn't designed to obscure a distinction; instead, it describes a situation in which changes in technology have blurred what was previously a fairly sharp distinction.

I find it useful to be reminded that many web applications have a substantial (or perhaps any) downloaded component, and that that may not be free. The rest of the article is very helpful and very reasonably worded; why did this slip through?

I've been a huge fan of Firefox since it was Firebird 0.7, or whatever the heck it used to be called. On Windows, IE is just a gaping security hole, and since I switched to a Linux desktop, not even an option. The web that I browse bears little relation to the one I see casual computer users interact with: I see hardly any advertisements, and monkeys don't jump around on my screen singing songs about products I can buy. But the last year or so, and the last few months has been taking a toll on my satisfaction with Firefox:

• Firefox 3.0.6 and later crashes abruptly, without warning or obvious trigger, multiple times per day. On the bright side, it crashes instantaneously and restarts in well under a minute, remembering all of my tabs, so I usually don't lose anything more than a minute. I just lose my sense of security and comfort, and get distracted into tangents like this post.
• The file open dialog is hideous. I shudder every time I have to attach a file, because this window comes up and has broken auto-complete. I'm not a big tree-clicking person; I prefer to just type the full path where my file is. But the Firefox file open dialog (some people say it's actually the GTK dialog, but even the bug reports are fuzzy) waits for you to type a few letters, then guesses what you are going to type, then erases the letters you are still typing, so that you have a path that's simply nonsense letters. It's actually sensitive to how fast you type. This is the sort of bug that I thought was left in the last millenium, but it seems to have resurfaced in Firefox over the last year and it sucks.
• Did I mention that it's crashing multiple times per day? I even down-graded back to 3.0.3, and it's still crashing. WTF?
• The certificate dialog. When you go to a website that has encryption, but didn't pay the encryption cartel for the privilege, you have to navigate through a series of obscure dialogs that, if you know what you are doing, are a silly waste of time, and if you don't, are very unlikely to help you. You have to click, in sequence, like this:
  1. a hyperlink: "Or you can add an exception ..."
  2. a button: "Add Exception ..." (your other option is "Get me out of here!", which is an utterly bad choice of words: please do not put informal idioms in the user interface of globally used software. Some of us won't understand it, and the rest will just roll our eyes every fracking time we have to see it.)
  3. a button: "Get Certificate"
  4. a button: "Confirm Security Exception"
• Now that it's crashing so often, I am noticing that the first time I input a URL in the address bar, the whole browser freezes for five or ten seconds while it does some eldritch ritual. Awesome.

Firefox is amazing in many ways, but there seem to have been a lot of mis-steps and poor decisions piling up recently.

I haven't had a TV in many years, but I do watch a few shows. Currently I'm watching BSG and How I Met Your Mother. Both of these shows happen to follow the current trend in TV: story arcs spanning multiple episodes and even seasons, combined with slow exploration of mysteries. With only one episode of BSG to go, and absolutely no trend of revelation in HIMYM, the obvious resolution just occured to me: Ted is going to marry Starbuck.

A few interesting articles went by recently about tax rates. This New York Times blog post points out an ABC News article in which it appears that the reporter didn't understand the difference between marginal and average tax rates. This doesn't require higher math, but I found it tricky enough that when I suddenly understood the difference, it was definitely an a-ha moment. The articles explain the issue, but perhaps a visual aid would help more people reach the a-ha.

In the graph below, the X axis is your pre-tax income, also called your gross income. The Y axis is after-tax income, or "take-home pay". In order to focus only on the difference between marginal and average taxes, I'm ignoring deductibles, allowances, payroll tax, FICA, joint filing, local taxes, and so forth. The only tax we examine is Federal Income Tax. I used the 2007 tax rates. The first chart goes up to $40,000 in income:

The red line is how much income you take home if there is no income tax. Unsurprisingly, pre-tax and after-tax income are the same if there is no tax. You make $20,000, you take home $20,000.

The yellow line is what you take home if you pay an average tax. I'm using the Wikipedia terminology here. In this example, if you earn $7825 or less, you pay 10%. If you earn between $7825 but less than $31,850, you pay %15. I.e., you are in the 15% bracket. More than 31,850, you pay 25%. I plotted after-tax pay at $1000 increments, and you can see clearly that when you go from $31,000 to $32,000, the amount of money you take home actually drops, because you went into a higher tax bracket. If you were making $31,800 and you accepted a $100 raise, your take-home pay would drop by thousands of dollars. It wouldn't make sense to take a raise unless it was many thousands of dollars, so that you would make up for the higher tax rate. This is how many people believe the US income tax system works. This is wrong.

The blue line shows a marginal tax. You pay 10% on the first $7825. Then you pay 15% on the amount between $7825 and $31,850. Then you pay 25% on the amount above $31,850. When you get a raise from $31,000 to $32,000, taking you into the next marginal tax bracket, your take-home pay increases. The only portion of your salary you are paying 25% tax on is the amount over $31,850, i.e., the last $150 dollars. This is the actual federal income tax system in the United States.

The tax brackets don't stop at 25% and $31,850. The chart below shows all of the brackets in the 2007 tax system, including the highest bracket, 35%, which starts at $349,700.

As you can see from the smooth curve of the blue line, it is always in the taxpayer's interest to earn more money. Each additional dollar of income always leads to increased after-tax income. Some people argue that the higher tax rates at the top reduce the incentive to earn more, because that 349,701st dollar is only worth 65 cents instead of 68 or 70 or 75, but I don't want to address that argument here. The point I'm trying to make is that the US has a marginal tax rate system, and earning an additional dollar never reduces after-tax income.

For analysis that moves beyond basic innumeracy, see this article at FiveThirtyEight and the stories it links to. The point there is that focusing on the rates and ignoring the bracket levels is also a mistake, and that maybe we should re-introduce higher rates for much higher income levels.