chunkbot

This statement only reflects a poor understanding of what it means to be "based on Prolog". I know Prolog well, and use Erlang every day, and I can tell you that they share nothing in common beyond the most superficial of commonalities. http://stackoverflow.com/questions/3542891/erlang-programmin...

When someone says $300 million dollar company, it doesn't mean a company with $300 million dollars of revenue or profit per year. Generally, it means the valuation (based on any number of measures, ie. acquisition price, last traded price x number of shares outstanding, founders pulling a figure out of thin air, etc.) of the company, so revenue can be anywhere from $0 to $300+ million (generally around single-digit multiples), and profit can even be negative.

Would FSF really use a "dislike" button if it were available?

Good stuff. Compared to the more "popular" generic behaviours like gen_server, gen_fsm doesn't get a lot of attention. It's nice to see it added to your tutorial with this level of detail.

I'm pretty sure that the majority of people here use an Apple computer.

Everything old is new again.

"It seemed to frustrate Appelbaum, who has a vivid and compassionate way of talking for a programmer, that Tor seems forbidding. [emphasis added]"

The problem is that people starting to raise families usually move out of NYC and into NJ suburbs with much better private and public schools (and a bigger, better house).

Delicious lived up to its founders' potential when it was acquired for millions.

C# the language is exciting. C# the platform, however...

I used to think the same thing about people here on HN, but I realized that a lot of accounts are tied to peoples' real names (and professional lives). So there's a filter that keeps people from discussing anything outside of politically-correct, nice-guy stereotypical talking points. But if you went "off-the-record", you'd probably find quite a few men open up about things like PUA techniques.

You're surprised? Why?

How do the Chinese do business? And how is it different from how U.S. Tech companies do business?

My favorite landing page is the one for Buckyballs (http://www.getbuckyballs.com/). Thanks to it, I'm sure they moved a lot of product on the day Google doodled "buckyball".

I don't think they're orthogonal; they're probably positively correlated.

I'm skeptical when anyone says a upcoming product is going to make running a fault-tolerant scalable environment "trivial".

An API for a third-party to query app-store data (things like Genre, Position, Name, Icon, Screens, Free/Paid, Reviews, Popularity, etc.).

Update:

Turns out Apple has this (iTunes & App Store Affiliate Tools), including the revenue share with affiliate networks.

Do any of the mobile app stores have an API?

Developers could come up with better app stores that drive more sales. Bonus points if they get a small percentage of every sale that flows through their app-store app. “Let a thousand flowers bloom and we'll see what flourishes.”

Haven't seen this before... I think this "Google" thing is going to be big.

I'm just surprised that a search engine I've never heard of has 800 servers, each with 64GB of RAM and 8TB across 8 disk.

(OK, I've heard of them, but now I'm interested enough to take a serious look!)

These are hard problems to solve, even approximately, and nothing we're doing is original. Like quantitative finance, a lot of the original research comes from outside the field (ie. "quasi-clique detection in protein-protein interaction networks"). For those that are interested, there's a lot of information in recent public or easily-available research papers. A weekend spent programming a good idea from a paper is well worth it.

For better or worse, we don't use any graph-specific heuristics beyond general assumptions of the graph's structure (which may or may not be correct, but that's another story...). We're dealing with a massive sparse graph whose vertex set, but not edge set, fits in memory (RAM). Embarrassingly, the edges are stored in a database in MySQL; we might be running the world's largest graph on MySQL, but I think we get better performance from MySQL than we could from any other products. Needless to say, we're not using anything relational. Like a lot of graphs (the Web, social networks, etc.), the number of edges is several orders of magnitude larger than the size of the vertex set, so we plan accordingly.

Perhaps my use of the term "partition" above is incorrect; these aren't perfect partitions (they don't exist, except in theory), but are what's known as "quasi-cliques" (the field of graph theory is rife with jargon).

We're happy with what we've come up with; contrary to most, the limits we face are mostly due to limited storage capacity rather than limited computation time or network capacity, which is essential for handling growth.

Sorry if all of this has been too general; there's not too much I can say specifically without revealing our "secret sauce". :D

I don't know about that. While being a good hacker probably correlates with being a good engineer, I think that the focus on "hacking" excludes a lot of very capable people. I know from personal hiring experience that a lot of qualified people are put off by the emphasis on being a hacker.

To expand on this, I don't think winning a multi-round programming competition (that is, successfully solving multiple algorithm-based problem statements under a time constraint with the highest accuracy and speed) is the best indicator of a good hire.

We've experimented with this, and we've found that with "social network" graphs (that is, scale-free networks whose degree distribution follows a power law) it's possible to determine "good" partitions on which to split the nodes. "Communication" (that is, algorithms that run across edges) within a partition is fast, and message-passing between partitions is minimized.

This would make for a fun A/B test: "$15 + $5 extra" vs. "$20 - $5 discount".

You might collect some evidence that people like getting a deal more than getting hit with an extra cost.

I upvoted you, but I was mostly just quoting the jwz article from 1998.

Please. IF you're using Chrome OS, you're "trusting" Google. so encryption that's impossible to disable is ok.

The existence of a garbage collector doesn't mean that the programmer has to play a totally hands-off role in the allocation of objects; as with any coding, there are usually a few bottlenecks that deserve special care. A system with a good GC will provide opportunities to tune; for example, to make hints to the GC about lifetime and locality and so on.

http://www.jwz.org/doc/gc.html

I treat memory as a resource too precious to be left to the mercy of programmers. In any application, a good garbage collector can be more efficient than malloc/free or new/delete.

http://www.jwz.org/doc/gc.html

How does Sony feel about subsidizing the cost of these things?