redxaxder

Number Theorists use a different definition of "almost all" from the one you see in measure theory. They use the limit version shown in the video.

A society with a massive surveillance dragnet that everyone has access to is not necessarily totally transparent.

Public access to the dragnet data would make it easier to evade the dragnet.

Bitcoin is also unusual for a collectible in how much it can be subdivided.

If we want the captcha to be accessible outside the target demographic, we should scrap the whole chess thing as well.

It bothers me that in some months, the 27th is 3 days from the end of the month, but in some months, the 27th is 4 days from the end. Could this mess up the graph?

The 27th appears in every month, but the time remaining to meet your (possible) quota would be different for different months.

If there really is a smaller barrier to developing for Ouya than other consoles, there will be Ouya exclusives.

In the US, judges rewrite the law with every ruling. Since the precedent established by a case is used in future decisions, the "real" law (the one which your actions will be compared against in court) is the base law passed in congress plus all of the precedent that has been piled on top of it.

It seems like most of the commenters here are treating the evidence of Google's wrongdoing as conclusive. It isn't. A comment on the post by Richard Wooding demonstrated how Joe Random can make requests from a Google IP address:

http://webcache.googleusercontent.com/search?sclient=psy-ab&...

By rescaling the numbers, the two questions can be turned into:

- "Would you pay $900 for a 90% chance to win $1000?"

- "Would you pay $100 for a 10% chance to win $1000?"

So the distribution of results really is different. It's not just a phrasing trick. I would still say no to the first and yes to the second. I like positive outliers more than negative ones. This doesn't seem irrational to me.

Every monad solves exactly this problem.

Between refreshes, the number of signatures needed jumped from 3 to 16. I wonder what happened.

This power is reserved for those secure enough against legal backlash to perjure themselves. Just like DMCA takedown notices.

Copyright holders with lots of resources can issue as many bogus DMCA notices as they like, safe in the knowledge that they can fight out or settle the (unlikely) lawsuit if it comes. Most people issuing fake takedown requests would expose themselves to a lot more risk.

This was written as a series of blog posts before being compiled into a book. The original is here:

http://youarenotsosmart.com/2010/02/24/fine-wine/

At least two differences between software and music:

- no malware

- possible online updates or other interactions

As long as these differences persist, your distribution method can easily add more value than the pirate copy.

No more than keeping your encryption keys a secret is security through obscurity.

Even if you're transparent about your overall scheme, you need to keep some execution details hidden.

When picking a password, you don't just care about the entropy. You also care how far down the password guessing order it is.

People who want to guess a password don't just brute force at random. They use a guessing order that goes through more common classes of password first. So if correct horse battery staple becomes a popular password scheme, these will end up attacked before other password schemes. (See http://www.schneier.com/essay-148.html)

Unless you're going to use a password safe full of nasty passwords, you should pick your passwords using an unpopular method.

Finding the convex hull of a collection of points is a studied problem, so you can use an existing solution:

http://en.wikipedia.org/wiki/Convex_hull_algorithms

Alternatively, willvarfar's approach is faster for large n, but the polygon won't be as pretty.

To implement a usable if, you need to tie it to some form of logical branching. Haskell can tie it to pattern matches. C compiles them to conditional jumps in asm. The asm jump instructions are implemented in hardware. The meat of the if definition lies in how you tie it to whatever form of conditional behavior your language has. This is missing from the article.

Smalltalk (presumably) has a form of logical branching in it somewhere. To finish the if, just pick one.

For a satisfying conclusion, you want to use one that doesn't already look like an if statement.

  equals: other
  ^ PTrue new

Will it take advantage of the ifTrue function you're building? Will it resort to some nasty arithmetic? Will it call some baked-in feature of smalltalk? Without an answer here, the "implementation" of if is just an illusion.

Rob Walling has written about this a bit.

"The 5 Minute Guide To Cheap Startup Advertising"

http://onstartups.com/tabid/3339/bid/43774/The-5-Minute-Guid...

3 part series on startup marketing, although this is more about retention than advertising:

http://www.softwarebyrob.com/2009/12/15/startup-marketing-mi...

http://www.softwarebyrob.com/2009/12/09/why-focusing-on-traf...

http://www.softwarebyrob.com/2010/01/12/startup-marketing-mi...

People haven't started paying $2000 for it yet. It may still grow.

Project Wonderful claims they make advertising awesome.

https://www.projectwonderful.com/

> I would guess most bitcoiners are mathematically inclined people, and not super into pure gambling.

Being mathematically inclined doesn't prevent someone from being into gambling. They are just less likely to treat gambling as a method of financial planning rather than a game.

The risk involved in gambling gives the game more emotional impact. This makes it fun, and (to some people) worth paying for as a form of entertainment.

More info on client SSL certificates:

http://www.gnegg.ch/2008/05/why-is-nobody-using-ssl-client-c...

Euclid's algorithm doesn't make as much sense until you see it done at least once. It's a shame there's no example. Finding the GCD of 85 and 221:

  221 = 2 * 85 + 51

  85 = 51 + 34

  51 = 34 + 17

  34 = 2 * 17 + 0

  51 = 34 + 17, so 51 = (2 * 17) + 17 = 3 * 17

  85 = 51 + 34 = (3 * 17) + (2 * 17)

  85 = 5 * 17

  221 = 2 * (5 * 17) + (3 * 17) 

  221 = 13 * 17

Bonus: This also gives us a way to write 17 as an integral combination of 221 and 85. We also get this by rearranging the parts:

  17 = 51 - 34

  34 = 85 - 51

  51 = 221 - 2 * 85

  17 = 51 - (85 - 51)

  17 = 2 * 51 - 85

  17 = 2 * (221 - 2 * 85) - 85

  17 = 2 * 221 - 3 * 85

> "I am not convinced it is that useful to have explicitly targeted audiences for advertisement. It may be better to understand and interpret an ad campaign, but generally, what works best when doing statistical prediction is not what is the most intuitive."

If you've collected marketing data that tells you your conversion rate among pet owners is 2% when your conversion rate among the general public is 1%, then you know that every ad you show to pet owners is going to have twice the impact on your bottom line.

Ad targeting enables you to collect this kind of data and to exploit it once you have it. If you find the right audience, your advertising dollars can go more than twice as far.

It's pretty neat that category theory is seeping into more programming languages. Even if Haskell doesn't survive, the functors and monads might live on.

As another graduate student who teaches, I can confirm jseliger's claim. Teaching difficulty drops of insanely fast. Don't think of it as making the task automatic, but rather having familiar territory off of which to base the things you say.

After one semester teaching a class, you discover what ideas the students tend to find most confusing and what types of errors are the most common.

Having leftover notes from the last time you taught saves you you most of the planning work, too.

((Buffalo buffalo) (Buffalo buffalo) buffalo) buffalo (Buffalo buffalo).