Forgot your password?
typodupeerror

Comment: Mysterious "Aurora" attack not so mysterious. (Score 1) 49

by Animats (#47417787) Attached to: DHS Mistakenly Releases 840 Pages of Critical Infrastructure Documents

There's nothing mysterious about this. The problem is that if someone gets control of circuit breakers for large rotating equipment, they may be able to disconnect it, let it get out of sync, and reconnect it. This causes huge stresses on motor and generator windings and may damage larger equipment. This is a classic problem in AC electrical systems. A more technical analysis of the Aurora vulnerability is here.

The attack involves taking over control of a power breaker in the transmission system, one that isn't protected by a device that checks for an in-phase condition. Breakers that are intended to be used during synchronization (such as the ones nearest generators) have such protections, but not all breakers do.

Protective relaying in power systems is complicated, because big transient events occur now and then. A lightning strike is a normal event in transmission systems. The system can tolerate many disruptive events, and you don't want to shut everything down and go to full blackout because the fault detection is overly sensitive. A big inductive load joining the grid looks much like an Aurora attack for the first few cycle or two.

There's a problem with someone reprogramming the setpoints on protective relays. This is the classic "let's make it remotely updatable" problem. It's so much easier today to make things remotely updatable than to send someone to adjust a setting. The Aurora attack requires some of this. There's a lot to be said for hard-wired limits that can't be updated remotely, such as "reclosing beyond 20 degrees of phase error is not allowed, no matter what parameters are downloaded."

Comment: Web programming sucks. (Score 1) 504

by Animats (#47417227) Attached to: Normal Humans Effectively Excluded From Developing Software

Ignoring the racist whining, he has a point. Web programming really sucks. Even web design sucks.

HTML started as a straightforward declarative layout language. Remember Dreamweaver? Macromedia's WYSIWYG editor for web pages. It was like using a word processor. You laid out a page, and it generated the page in HTML. It understood HTML, and you could read the page back in and edit it. Very straightforward. You didn't even have to look at the HTML. Back then, Netscape Navigator came with an HTML editor, too.

Then came CSS. DIV with float and clear as a primary formatting tool (a 1D concept and a huge step backwards from 2D tables), Javascript to patch the formatting problems of CSS, absolute positioning, Javascript to manage absolute positioning... The reaction to this mess was to layer "content management systems" on top of HTML, introducing another level of complexity and security holes. (Wordpress template attacks...)

It's as bad, if not worse, on the back end. No need to go into the details.

All this is being dumped on programmers, with the demand for "full-stack developers" who understand all the layers. Cheap full-stack developers. Usually for rather banal web sites.

Not only is this stuff unreasonably hard, it's boring. It's a turn-off for anyone with a life.

Comment: Being a quant in the early years. (Score 4, Interesting) 87

by Animats (#47413241) Attached to: The Billionaire Mathematician

His fund has an impressive trading record. He had the big advantage of starting early, in 1982, when almost nobody was doing automated trading or using advanced statistical methods. Their best years were 1982-1999. Now everybody grinds on vast amounts of data, and it's much tougher to find an edge. Performance for the last few years has been very poor, below the S&P 500. That's before fees.

The fees on his funds are insane. 5% of capital each year, and 45% of profits. Most hedge funds charge 2% and 20%, and even that's starting to slip due to competitive pressure.

Simons retired in 2009. You have to know when to quit.

Comment: Re:What happened to Scheme? (Score 1) 393

by Animats (#47411269) Attached to: Python Bumps Off Java As Top Learning Language

Abelson and Sussman is a delightful book for programming theorists. Scheme is a big improvement over Common LISP. Learning Scheme from Abelson and Sussman is straightforward for people who can get into MIT.

This is not most of the programming population. As someone else pointed out, programming today is mostly the creation of glue code to tie together a number of (usually buggy) components. Neither the webcrap crowd nor the appcrap crowd needs Scheme. In fact, if you have that strong a theoretical background, you tend to overdesign simple programs.

Comment: Good idea (Score 5, Insightful) 393

by Animats (#47410267) Attached to: Python Bumps Off Java As Top Learning Language

Python isn't a bad first language. It has all the important advanced concepts - objects, dictionaries, closures, and threads. The syntax is reasonable. Some people are bothered by the forced indentation, but for new programmers, it will seem natural.

Most of the problems with Python are performance related. They come from obscure features of the language, such as the ability to do "getattr" and "setattr" on almost anything, including objects running in another thread. So everything has to be a dictionary. (This is sometimes called the Guido von Rossum Memorial Boat Anchor.) PyPy is struggling hard to overcome that, with some success. (The optimization approach is "oh, no, program did Obscure Awful Thing which could invalidate running code" - abandon compiled JIT code, shift to backup interpreter, flush JIT code cache, execute Obscure Awful Thing, wait for control to leave area of Obscure Awful Thing while in backup interpreter, rerun JIT compiler, resume running compiled code.)

Comment: The frustrations of AI. (Score 1) 551

AI as a field suffers from the delusion that we're one breakthrough away from strong AI. There were people saying that at Stanford in the mid-1980s when I was there, just as the "expert system" hype was failing. There is progress; the current generation of machine learning can do some things quite well. But it's not leading to strong AI Real Soon Now. We can't even do dog-level AI. Or mouse-level AI. Insect-level AI, yes. (It was bacteria-level AI in the 1980s. There is progress.)

More likely, we'll get robots that can sort of deal with the real world, and they'll be improved over time. (It's embarrassing how lame robotics really is, after 50 years of R&D. DARPA is trying to kick some ass with the DARPA Humanoid Challenge to get machines that can do something useful other than work an assembly line.) We'll get programs which can deal with most business problems ("Microsoft Middle Manager 3.0"), and they'll be improved over time.

Hardware is not the problem. If it were, we'd have things that were very smart, but very slow. Then someone would rent enough Amazon AWS instances to make them fast.

Comment: Earnings reports are in XML now. (Score 4, Interesting) 29

by Animats (#47389353) Attached to: Algorithm-Generated Articles Won't Kill the Journalism Star

The SEC started requring companies to file their earnings reports in the Extensible Business Reporting Language a few years ago. At first, it was only for big companies; now it's everybody. The SEC displays this info in a standard format on line. Here are the latest earnings for DICE Holdings, Slashdot's parent. Here's the raw XML behind that data. Turning that into verbiage isn't that hard.

I've been doing this for years at Downside.com, extracting the raw data from the human-readable text. This is now obsolete, but it's still running. Here's the same DICE financial statement as processed by Downside. That's Perl code that's been running for 15 years now. When it started, nobody was doing that. Now that everybody in finance has that data, it's probably time to retire Downside's old extraction engine.

Comment: Another materials article (Score 3, Insightful) 33

by Animats (#47381759) Attached to: Researchers Create Walking, Muscle-Powered Biobots

It's another materials article. They do not have a "walking robot". They have one piece of synthetic muscle fiber hooked up to some supports. If they hook up an oscillator to the power, it jerks along. There are other artificial muscle technologies. This new one is supposedly powered by the chemical solution in which it operates, not by the electric field that triggers it. That's new. But if it's chemically powered, there must be waste products from the reaction that have to be flushed out. You need a whole circulatory system for the thing.

"An organization dries up if you don't challenge it with growth." -- Mark Shepherd, former President and CEO of Texas Instruments

Working...