I am at a complete loss to understand how a project that is completely open source (https://fuchsia.googlesource.com/) can be considered in any way "mysterious."

I recently tried to explain to my mother how the programs I run can be "free." It's hard, from the inside, to appreciate how much what we do looks like magic. I tried to explain that web pages are just like Sudoko: Sudoku and CSS are both constraint solution problems; we just have a language for handling them. All the tech for word processors, browsers, spreadsheets, and databases are, well, "common knowledge" to me. CSS is a constraint algorithm. Word processors are gap buffers and Aho-Corasick algorithms. Spreadsheets are reactive graph traversals in a functional setting. Databases are B+-trees. To me, that's just... stuff I know. To my mom, that knowledge must not only be esoteric, but so esoteric that it must be privileged and expensive. Trying to explain that it's first year textbook stuff no more privileged or protected than long division just goes right past her. So trying to explain that "some guy" wrote a word processor because he thought that would be more fun than model airplanes or bowling league just kinda blows her mind. Oh yeah: all cryptography is nothing but long division; it's just done with numbers so big only computers can do it in a reasonable amount of time. (Not to diss bowling leagues. I was in one once. They're fun!)

The 1996 Penguin Hutchinson Encyclopedia Library (PHRL96). I keep that running in a Wine-managed desktop window more or less constantly; I've tried on-line encyclopedias like Artha and Panlexicon and even Wordnet, and the thesaurus in PHRL96 is still the best one I own. Also: Half-Life. The original. Recently re-played it, and it works wonderfully.

Adam Smith, the "father of economics" and one of the original theorists of capitalism, believed that capitalism worked because each participant in the marketplace had an approximately equal capacity with respect to other participants to understand the value which he was exchanging with others. Some people are more clever, or have better memories, or are simply more industrious, but on the whole we are all human beings, and our ability to know more than another has an upper bound.

That's not true of machine learning. There is no upper bound. Under surveillance capitalism, there is no limit to what the large companies can know about what you know, can monitor what you do, and can predict what you want. And as long as we remain human, that upper bound on large, corporate control of human beings will only get greater.

Sounds like she's well on her way to being a full-stackoverflow developer. I know a few of those. I work with a few. Some of them are surprisingly good at their job, as long as their jobs are producing working web-code on a deadline. It's not creative or clever. All of the actual mathematical work is done for them. They're fine people who would be utterly and completely lost if Google and Stack Overflow weren't there to point them in the right direction.

I read the paper. It didn't need to be a programming language. The paper has three separate parts: a mechanism for adapting standalone robots of multiple designs and with different roles, into a complex, goal-achieving swarm. First, it does this by coordinating tasks and flocking behavior through "virtual stigmergy." Stigmergy is how ants and bees coordinate: they leave traces in the environment that indicate task, distance and direction to task target, and when those traces pass a threshold some or all of the swarm *acts*. It's a cool idea, although flocking and stigmergy can be found in video games as old as the original Pac-Man and Rip-Off. Second, they expose the virtual stigmergy and awareness of swarm membership and swarm member locations through an interface (hardware and software) that sits on top of the usual robot command & control. They've written a VM that handles the real-time data management associated with these new "awarenesses." Third, they write an independent language targeted at their VM, with native access to the swarm(), member(), and stigmergy() objects. The language looks like Javascript. The authors aren't great language writers; they even comment about "the jargon of object-oriented programming." There was another paper: "Micro Virtual Machines as Solid Foundation for Language Development," that they would have benefited much from reading. If they'd targeted their VM to *just* handle the hardware, specialty concurrency issues, and garbage collection, and left everything else up to independent language writers, they wouldn't be asking people to program their robots in two different languages (one for swarm management, one for robot management), and the language people could have created highly optimizing compilers, libraries and languages targeted at doing virtual stigmergy well.

No kidding. I'm glad we didn't. It means I can look at myself in the mirror. Career-wise, I've done okay without it. But it would have been a completely legal patent through which CI$ would have raked in millions and mililons of dollars. And, as far as I can determine, it would have been completely legal. There was no MySQL, no Postgres; OraPerl had *just* been released and was barely stable on SunOS, and there were no known instances of a CGI / OraPerl gateway on the Internet until Pacific Power & Light asked us if it was possible to connect their consumer-oriented energy savings database to that new thing called "the world wide web."

In 1993, I failed to file the US Patent on "A means of accessing a relational database via the Internet." If we'd known we could do it, CompuServe might still be around.

Intel i5-2500K, ASUS P8P67, 4GB DDR3-1600, Seagate Barracude 1TB drive, Nvidia GTX-480, Fractal Design R3 Gamer Edition. This is basically the 2011 Ars Technica Hot Rod, but with the Nvidia swapped for the recommended Radeon; at the time, Nvidia support for Linux was much better. Also, I couldn't find the R3 "standard" case, but Fry's had the "gamer" version, which was probably a very good idea; it's chunky and slightly noisier, and it has a lot of deep blue LEDs in it, but I have never seen this thing get much above 110F even under incredibly heavy gaming or transcoding loads. I really should throw more RAM into it.