bcater

Thanks for your suggestions, all. Our traffic roughly doubled, which didn't cause any strain on our infrastructure.

For the curious, the site is http://betterific.com, which was linked from http://facebook.com/arbys, marking our first big partnership. After endless market surveys about what your customers want, Betterific tells you why.

Start one. Ordinary people do it all the time.

I recommend less text shadow.

I have a similar laptop: 32-bit, 1Gb RAM, 1.6Ghz atom, purchased years ago on the cheap. I run the latest Ubuntu desktop edition and am able to run everything that I need (mostly Rails + remote services over SSH tunnels). I can even stream Hulu, though the video does sometimes lag.

I don't have personal experience with any other distros, but a friend of mine has an equally antiquated laptop running the latest Fedora which he claims is better than an abacus. YMMV.

Ping me - I've got some ideas. [email protected]

For those (like me) who didn't already know, the FetLife homepage appears NSFW.

FWIW, I worked at an online advertising company that ran a whole division on EC2 machines - at the peak, nearly 100 instances. This was never a problem for us because we were judicious in our use of the machines (minify your content!) and we were disciplined in our deployments and management. If you go that route, just be sure the learn and follow best practices - you'll save yourself quite a lot of time.

I'd recommend "The Mythical Man-Month" by Fred Brooks.

You're completely right. Thanks for the find!

Taking ideas from the beginning of "Keep Your Identity Small," I'd like to read something about art and / or artists. I choose that topic because, on one hand, I can say without justification that I prefer one artist's work over another's work, yet there must be some justification to say that one artist is obviously better than another artist. I'd like to see pg's treatment of a topic that is based in both opinion and fact (and one in which he has some expert knowledge).

Ditto on the recommendation. It's one of the few free pieces of software that does something that's important to me really well.

I started using it this morning. After my session auto-expired, I can no longer log back in. It seemed useful until that happened, but lack of reliability is a deal-breaker.

Riddled with grammatical errors, this piece makes me chuckle.

After 1 million trials with the above method I get: avg: 499.00 sigma: 1115.04 [499.00 - 3.35,499.00 + 3.35]

You're exactly right. Whoops!

This is all scribbled in a notebook, so there's a good chance that it's wrong, but bear with me on the arithmetic.

As a first guess, let's agree that in expectation you earn $0.51 each time you flip the coin, so it should take you about 20 tries to reach $10. Let's do something better, though.

EDIT: The previous paragraph is totally wrong. Thanks, tome :)

Let's build a confidence interval with alpha = 0.01 so that (1 - alpha) = 0.99. First, we'll need some trials. For that, I wrote a program that flipped a weighted coin and played the game until it reached $10 using the rules that you described. I recorded the number of coin flips required in each of 15 trials:

298, 84, 268, 2712, 110, 66, 42, 128, 84, 48, 280, 80, 64, 42, 234

We'll need the sample mean, X_bar = 302.

Now, we'll compute the Z-score so that we can build an interval in which the true mean (mu) lies with 99% probability:

P(-z <= Z <= z) = 0.99

We know that Z = (302 - mu) / (sigma / sqrt(n)), where sigma (the standard deviation) = 650 and sqrt(n) = 4. I'm rounding. Therefore, Z = (302 - mu) / 168.

Now, let's look at the cumulative distribution function Phi(z) and note that if Phi(z) = 1 - (alpha / 2) = 0.995, then Phi(z) ~= 0.997, the approximate cutoff for the 3rd standard deviation. Thus, z ~= 3.

Thus, we have that P(-3 <= (X_bar - mu) / (sigma / sqrt(n)) <= 3) = 0.99, so P(X_bar - 504 <= mu <= X_bar + 504) = 0.99. Therefore, I am 99% confident that the true mean, mu, lies on [X_bar - 504, X_bar + 504] = [302 - 504, 302 + 504].

That's a really wide range, and seemingly completely unhelpful for the purposes of betting. More sample trials would teach us more and lead us to a smaller interval since we expect that within some large number of trials we will converge on mu.

CNN says that he is in the hospital:

http://www.cnn.com/2009/SHOWBIZ/Music/06/25/michael.jackson/...

Interesting that they use "type" as a field for the Format class:

class Format(models.Model): type = models.CharField(choices=...

type() is a Python function, so it doesn't make sense to use "type" as a field name. I found out the hard way that this causes bugs.

Authorize.net - http://authorize.net/

At the time that I took 6.001 (the MIT version of the SICP course), PLT Scheme lacked several libraries that the book requires. Since so many students wanted to use PLT, the staff put together the extra modules for us, but I'm not sure if PLT is now "SICP-compliant" or not. Check with someone who works on PLT (they're very nice) to find out.

For what it's worth, I really enjoyed SICP, though I still can't say that I understand every line in it.

I was discussing building an internal payment management system over lunch today. Do you have an API that we could use instead?

As others have said, kudos on the design.

What a cool company! JamLegend, that is.

I'm not sure what your use case is, but we've had much better luck with generating LaTeX code that we compile to PDF (the trivial step) than with doing HTML to PDF.

Remarkable.

Knuth all the way down at 16? Tisk.