Notice that the dominant factor is the 1600 strings to be matched. If you really care about performance, then you want to get rid of that factor.
True, but why did you launch directly into automata? A hash-based bloom filter k=1 which tests each word against a bit vector of 8KB (64kbit) in size (L1 friendly) will flag roughly one false positive out of every 40 words (p=1600/2^16) for further testing. On the first true positive, the algorithm terminates, so this does not contribute to asymptotic run time. Depending on your CPU, the bloom filter might cost up to twenty instructions per character processed including amortized detection of word boundaries.
Then you need to include additional processing triggered by 2% of words evaluated which can also be done with efficient hash methods (likely using larger non-L1 data structures). If you haven't banned any 1 or 2 character words, you can skip these completely. The second phase processing won't add up to much.
Note that you should use a randomized hash function so that there is no predictable worst-case message available to miffed DOSers if you think this is a task worth doing well. This costs you at initialization time (dynamic instead of static bitvector), but is free thereafter.
The easiest dodge for the person writing the message is to replace every space with an infrequent character such as 'z' if the words are hard to read written with no spaces at all. Or you can go full camel case. If the rule for what determines a word boundary becomes shifty, you're into full-blown spam filter hell. Of course, in such a country, shifty word boundaries could cost you your left one, so parse prudishly if you value your thumbs.
In hindsight, perhaps, this is all clear. At the time, would you have bet your house on the proposition of 386BSD remaining unscathed if the BSDi lawsuit had come to a different outcome? But wait, I have a reference.
Like the other groups, they started by adding the six missing files that Bill Jolitz had written for his 386/BSD release.... At the preliminary hearing for the injunction, BSDI contended that they were simply using the sources being freely distributed by the University of California plus six additional files. They were willing to discuss the content of any of the six added files, but did not believe that they should be held responsible for the files being distributed by the University of California [which 386BSD also used, one would think]. The judge agreed with BSDI's argument and told USL that they would have to restate their complaint based solely on the six files or he would dismiss it. Recognizing that they would have a hard time making a case from just the six files, USL decided to refile the suit against both BSDI and the University of California.
Yeah, totally clear how 386BSD was free and clear of the legal fog of war. And a huge debt owed by everyone to Marshall Kirk McKusick and friends who fought this battle on our behalf while Linux thrived under the legal radar.
In my own view, Linux had a crazy-making anthill culture, which appealed to many young coders with more energy than brains. But you know, I wouldn't bet against energy in retrospect. The annual ipchains rewrite boggled my mind. Not my cup of tea. An even crazier splinter group made hay with PHP, breaking just about every rule of thoughtfulness and elegance known to God and man. And look where that got them: pretty damn far.
I would personally, however, have jumped on the BSD wagon at the time had it been able to promote a coherent vision of life after lawsuit. What would be the balance be now if BSD had gathered twice as many elitist greybeards into the fold? I have a feeling it would have continued to lag in the department of crappy consumer product device drivers, compromising a major defection path from Windows 98. Greybeards don't do popularity worth a damn.
Debian zealots notwithstanding, Linux quickly became popular enough to become a willing host for binary blobs.
They still argue to the cause, but considering that the year 2011 saw unprecedented production of greenhouse gases (far exceeding even the worst case scenarios), it should now be clear to anyone who doesn't have a personal axe to grind that the climate is in the process of extraordinary change, and that the conditions we rely on to feed 7 billion people are about to get very dicey.
With seven billion people on the planet, and rapidly rising standards of living in China and other parts of Asia, with coal increasingly augmenting the petroleum plateau, I'd be shocked if every year didn't set a new record for CO2 emissions.
What would be unprecedented here is a global consensus among rich and poor, east and west in preventing business as usual from raising the bar on CO2 emissions every year for the next half century--which is exactly what would have happened had the earth's CO2 levels not been precisely balanced at a precarious tipping point as science presently tells the story. If we pull this off, we'll be manufacturing a political consensus out of whole cloth such as never before witnessed on this blue marble.
On another note, I don't get this beautification of scientific consensus as the second coming of fast food culture: science is, and always was, a slow food movement. It takes decades or centuries to reach secure conclusions concerning systems as complex as the earth's climate. I think this is a lot like a doctor who discovers a new disease model, then immediately proposes an extremely radical treatment of unknown severity and consequence.
Via Wikipedia:
Monsieur Homais is the town pharmacist. In one incident, he convinces Charles to perform corrective surgery on a young stable boy, afflicted with a club foot. During this era, correcting or eliminating a disability was a daring option and he may have considered this an opportunity to garner personal attention and praise. The operation is a disaster, and the stable boy is left with his leg amputated at the thigh.
Amputated at the thigh IIRC by another doctor who shows and takes responsibility. In the long run, these interventions become routine, and the consequences become understood and mitigated.
Is there any evidence that we can fundamentally shift the global economy away from fossil fuels on a radical program without incurring large and unknowable risks to geopolitical stability in doing so?
The paint is still wet on climate science. Be careful what you wish for. And don't write me off as a club foot surgery denier. The old day-glo Wired was my personal hot tub: I'm a card-carrying techno-optimist. Politically, however, I'm extremely wary about any combination of alacrity with wet paint. Apollo 13 was pretty much the historical high water mark on smooth sailing amidst a crash program instigated by handshakes among our political overlords.
The big difference from Bell's theorem is that in Bell's theorem, the quantum states are known to be entangled.
In every experiment I've read about, the entanglement is known by how the particles are created. But here's the question I never see answered: is it possible, given two particles you know nothing about, to prove the particles are not entangled?
If you can't prove any given pair of particles are not entangled, then perhaps entanglement is the natural state, and particles known to not be entangled (by some exogenous information about their creation or history) would be the exceptional case.
In cryptography, the entanglement of key bits and cipher bits quickly becomes so diffuse you can pretend it's not there, for almost all practical purposes.
If all particles became entangled at the time of the big bang, then maybe there really isn't such a thing as unentangled particles. Now, I don't know nearly enough about physics to know how this suggestion could be easily blown out of the water, but I know enough about philosophy to puzzle about why this very basic aspect of the paradox consistently escapes clear explanation.
Perhaps the only entanglement one can observe is the one where you know something ahead of time. Then to say that two particles are not entangled is more a statement about your baseline ignorance than it is about the particles.
Chernobyl and Fukishama have now both shown that nuclear incidents are ALWAYS worse then estimated...
I suppose what you are saying is that people ALWAYS exceed the speed limit unless the speed limit exceeds the reigning land speed record, but that such a speed limit could never be adopted by any social process even though Germany has in fact adopted something not entirely different.
When the sun finally goes red giant, I'm not entirely sure the damage to the planet from nuclear energy will actually be worse than estimated.
I'm pretty much 100% certain that somewhere between 1950 and 2100--as things are presently progressing--we'll gain enough engineering and political competence to make nuclear energy a safe alternative relative to any sane norm in these matters, if by then we still wish to pursue it, which is highly doubtful, but not impossible.
Evidence now appears overwhelming that 1970 fell short of the mark and that 2010 has yet to complete its homework assignment.
Note: you can actually see an impressive long march towards human political competence through a thousand year aperture.
I guess now we wait and see whom Apple will use this against?
What is "this" exactly? The fact that they now hold the patent, and might someday do something with it? Actual power obtained from this patent after surviving a legal challenge by a motivated and cash rich adversary?
If Apple didn't hold this patent someone else would use it against them.
This kind of circular reasoning is what tips the game theoretic balance against having the patent system in the first place.
I was pleased with the discussion about this kind of dynamic in Satz on Markets.
The old view that transactions take place between voluntary participants in the absence of negative externality is pretty much dead and buried given the complexity and interconnectedness of the modern economy.
The local utilities give out dividends. If you want steady income, that's what you buy. It's boring. It's low risk. It's slow/minuscule/nonexistent growth.
That's what Microsoft has become: the office work flow utility. Microsoft worked very hard to become such a utility, all the while distracting the audience with the other hand lip-synching the word "innovation" which as everyone knows is mainly MIA.
They will never describe themselves this way directly in any media that gossips to the DOJ. You just have to look at their stock performance and the people who invest. It's obvious.
If it looks like a utility through the moneyscope, it is a utility.
I've been reading since age 3 and read at 1500 wpm with 100% comprehension. I could have told them long ago that this is how I do it.
If you ask a group of people to self-assess for leadership skill, eighty percent report above average leadership skills. What seems to be happening is that each person defines leadership as heaps of whichever skill component he/she happens to possess, and not so much of the skill components he/she lacks. They are all telling the truth with respect to variable criteria.
So I'm wondering, does autism define "comprehend" the same way a non-autistic person does? I read about half that speed, and I don't feel limited by word recognition, but more by the multiple processes of figuring out where the author is coming from (or not, if the agenda is to apply lipstick to a mental vacuum). There are so many layers to discourse analysis it's hard to list them all.
"Comprehend" could mean retaining information points presented as fact. Or it could mean assigning the dribble of factoids into mental categories "pulled from ass", "brandishing urban legend", "regurgitated from recent popular news story", "manufactured in a pique of convenience", "seduced by right-thinking glean", "outright deception", etc. It's a lot of work when reading to man the airport scanner of psychological bogosity.
Furthermore, these assessments are fluid and tentative and require a large working reserve of rewritable storage. A model of the author as a reliable or unreliable human being is formed, if the assessment is to care enough to do the work.
This last effect is most obvious watching movies. I give weak passing grades to diverting films I couldn't care less about. If the movie gets just enough better that I start to care about what it might have been, that's when my harshest judgments are unleashed: I've entered into the punative "made me look" valley where I actually turn on the critical machinery--often to discover not entirely quickly enough that it was a false start. The subset of the movies that make me care and then reward the bother is where I start giving out decent scores.
Sometime when reading I turn the page, and go "ugh" inside and then feel the overwhelming urge to skip forward half a page or a whole page, or both pages. Then I go "how can you _know_ all this text is worthless in a tiny fraction of a second after the page flip?" So I go back and slog through it and sure enough, in the vast majority of cases, my instant assessment was right on the money.
Many of the long-winded essays linked from aldaily.com are particularly challenging in this regard. Some of those writers are talented enough to go on for page after page saying hardly anything at all, while defeating the immediate "this is vacuous crap" quick page-turn self defense. It appears that there is a high art to saying nothing in such an elaborate and convoluted way that busting the vacuity of the prose reduces me to my real-time reading speed.
I once read a piece, Kirkegaard I think, about chasing a bug around a desk with a pin while enduring immense boredom in the classroom. The humanities is where you learn to wield the pin, and make your reader perform as the bug. Not always, but fairly often. What to these people does the word "comprehend" actually mean?
As long as they can deal with the fact that I'm a 90 year old female that lives in 90210.
That's precisely what they want to know. Giving a different answer on every web form lowers your learned-helplessness score to almost zero. They won't want your business. You might not blindly pay charges you don't understand for services you didn't know you needed.
These bonuses are the banking equivalent of free drinks at Vegas. No one drinks free who isn't losing far more than the drinks are worth.
You think the smart people will simply head off to the smart banks that don't go in for this nonsense? Then how is it that the credit card companies have convinced retailers not to discount cash? People who pay cash miss out on all the great CC swag.
Ultimately they get you coming or going.
alternate life pursuits not entirely like living
on
2011 Geek IQ Test
·
· Score: 1
"Geek IQ" is a polite synonym for mastering alternate life pursuits not entirely like living. However, such a test ought to know the difference between nerd IQ and geek IQ.
The question about "Heroes" is the moral equivalent of the birthstone questions in the science category of the original Trivial Pursuit.
Every category had some bird food. In Sports, you just keep picking Bath Ruth or Mohamed Ali whenever they come up, you'll get one eventually. For a science ignoramus, the closest surrogate to low-hanging fruit is substituting astrology for astronomy.
Google is my friend, too, yet I would have been ever so grateful for the tiniest social grace of sparing me yet another Google result set.
I'm pretty sure Lewis Thomas in Lives of a Cell (or possibly Et Cetera, Et Cetera: Notes of a Word-Watcher) comments about the sad decline of the elder statesmen: he hasn't forgotten anything so much as piled it so deep in the attic he can't find it without a substantial jog.
For about five minutes a year ago I knew what Diaspora was. Then it went directly to the Lewis Thomas attic of things I can only possibly remember once reminded.
Hard to understand, I guess, when you're twenty two.
stop posting anything potentially incriminating or even marginally controversial under your real name
The world has become a kind and gentle place. Used to be the best advice was to sever your penis. Have you ever thought of signing up for Root Strikers? You're a natural. No, wait, I have it backwards: your advice is a gilded path to office in Washington.
Outside of this post, there sure seems to be a lot of people on this thread suffering from reality shock.
If assassins had a union, the union would terminate any member who chills business by snuffing out any prospective customer who shows up, unless everyone is pretty sure you kept your solicitation close to the vest.
I suppose it's like the priest who hits a hole in one on the golf course when he should have been preaching: Who can he tell? When you run with a crowd where you can reveal your murderous solicitations as insurance against the instant double-cross, we call it organized crime.
In any case, with this kind of intimidation, there's often a tipping point where there's too much to suppress and too much risk in doing so. When the heads pile up to the sky, it becomes a political issue.
On a podcast the other night, I heard that stern discipline of children promotes frequency and skill at lying. Some parents care enough to give their children a big headstart on a lucrative career path.
7 is at least two orders of magnitude better than the 9x series, and one above 2k and early XP... I'll agree.
The vast majority of blue screens I had with 2000 were hardware related: bad chip sets (Via KT7A), bad memory (many), bad capacitors (anything from Asia), bad disk drives (Fujitsu, IBM). Another class of blue screens pertained to immature device drivers. Video cards in particular were a fast moving niche, and developing for Windows 2000 was not yet a common expertise. I used 2000 for a long time because I never upgraded to XP. There was no significant stability problem.
Hardware maturity has also come a long ways since 2000. As a piece of software, I never felt I had stability problems with Windows 2000. You were somewhat at the mercy of the device drivers installed. Fail-safe sandboxing comes with a performance penalty that was less attractive on a 700MHz Pentium III with 512MB of DRAM. I think an order of magnitude estimate on instability of 2000 vs Windows 7 is gross inflation. If you can find a mainboard from 2003 that hasn't leaked its mojo, install Windows 7 and see how its stability rates.
I did have a fair number of blue screens under NT that were entirely NT's fault. The NT AppleTalk networking component was a major puntfest on my home network.
What keeps me on Windows: any single killer application I can't replicate under Linux.
What keeps me off Windows: imprisonment of my work product in proprietary vendor data formats.
One is a short term liability, the other is a long term liability. Migration continues. Speed of the migration would be enhanced if all geeks unite to eschew the nVidia binary blob.
Re:Next up : Compiler wars!
on
The Transistor Wars
·
· Score: 3, Informative
We could even see functional languages with layers of profiling metacode producing self modifying code that runs blazingly fast.
The computer capable of that level of introspection and inference would snort at your silly fashion bias toward functional languages. The main calling card of functional languages is to offset weakness in human cognition. The human brain struggles to convert a functional specification into an optimal state machine without dropping a stitch. Kasparov and others complain about computer chess precisely because a well-tested adversary never drops a stitch, or so rarely that chess programmers have a dozens of other things to worry about first.
You do realize that the primary virtue of a functional language is purity in the specification domain and that it offers no fundamental advance in the execution domain?
There are two halves to the specification domain: algorithmic correctness, and shaping the performance/resource envelope. Prolog is a fairly reasonable specification language for algorithmic correctness, but almost completely useless at shaping the performance/resource envelope. Ever seen a smart phone OS programmed in Prolog?
There needs to be a word for this particular cognitive bias. This is the cognitive bias that if there's enough food on the planet, no one should starve, neglecting only the distribution challenge modulo politics, history, culture, and human nature.
In a world dominated by programming languages optimized for algorithmic correctness, all our problems will miraculously go away, because all those potent algorithms will sort out the performance/resource envelope without further input of blood, sweat, and tears. Nice vision.
That day will arrive when I specify the desired solution as a shortest path and the computer responds, "no can do, but would you settle for nearly as good almost all of the time under modest stochasticity assumptions in the underlying graph in near real-time to the largest feasible problem size as practically bounded by performance bounds elsewhere in the application feature set as they presently exist for the targeted user base?"
And I will go, could you break that down into smaller pieces? I'm out of practice thinking that hard.
On the transistor topic, it's kind of stupid to neglect the power distribution tree. Idle execution units don't leak if the master valve is slammed shut. In future we can have a much larger set of execution units optimized for different tasks, and only use the one that's needed for a heavy lifting loop.
You're already seeing the shift to dark silicon with the introduction of the ARM A5 as a companion dog to a bigger OOO furnace. One or the other CPU is shut off completely at any given time. Hard to leak power that never arrives.
Propably won't for you, because it seems that you have already decided even trying it seriously:)
Those of us that have an open mind don't put such a low price on our precious gift. From what I've read, the ratio of show stoppers to features that might possibly make my life better (after a significant accommodation period) is about ten to one. My number one productivity feature right now is having nine desktops on two large displays. I've not read a single Unity review that says "expect a great experience with the dual-head multidesk".
Open mindedness is not carte blanche to compromise what already works to pursue a pot of gold at the other end of the usability rainbow. To make matters worse, I can't even see this particular rainbow, much less the pot of gold.
Ubuntu screwed this up. There should have been at least a full two year period where a feature complete Unity existed in parallel with the tried-and-true so that the open minded could pick their spot for making the switch.
If you've only got one ball in the air, and you don't mind this ball being adaptation to Unity, then I question whether you're an open minded person at heart, or just some chump who thrives on the reckless pursuit of novelty.
Zinga will still attract talent, but the contracts will have thicker prose about termination and vesting conditions, so it will be much harder to pull this stunt a second time. But perhaps at this stage of growth they are beyond that.
I watched Politics, Strategy, and Game Theory last night, which talks about Grim Trigger and the conditions under which, in iterated prisoner's dilemma, you care more about the future than defecting in the present moment.
It's a competent lecture with no great pizzazz.
Here's a fairly nice piece by an undergraduate I stumbled upon brushing up on Grim Trigger: Debunking the Prisoner(slashcode fuckup)s Dilemma on Robert Axelrod(slashcode fuckup)s Emergence of Cooperation among Egoists and why cooperation is a lot more common than the shallow analysis would have you believe.
I really wonder what payoff matrix he constructed to author that blog under no fixed identity. I found a Tweet referencing the site by the apparent author with one or two clues about his circumstance.
The punishment for Zynga in future iterations are employment contracts with a lot less room to wiggle if they screw up future hires. The reward in the present iteration is yanking back a substantial chunk of the entire company.
If the quiet vestors really aren't showing up and pulling their weight, it doesn't seem great to let them get away with that either. I don't think Zynga's presumption is that they can't fight this, but more like "the effort involved will be a shock to their lazy asses" so they are likely to settle without going ten rounds.
It's more appropriate to say "It's almost all US" at this point. China, having produced less CO2 in the past decades but now producing more, has only just started to catch up. It's got a long way to go.
I'm not sure assigning blame is the first order of business. I would slice things up a little differently. One of the main reasons that China's output is growing so rapidly is that the developed world has shifted manufacturing offshore. Much of the CO2 released from China is making the consumer goods we consume in America and Europe.
By the same token, the high levels of emissions in America and Europe correspond to the majority of technological progress over the last century, which the whole world shares. The low power cell phones in Africa were made possible by an immense technological investment elsewhere.
Which body organ do you blame? The more interesting analysis is whether we're using energy wisely. America especially uses a lot of energy rather stupidly. We build sprawling cities with massive road systems for single occupant vehicles instead of denser cities with excellent public transit. Some of those decisions might have panned out differently if gasoline prices had been a lot higher a lot sooner. Detroit didn't want it that way. Detroit wanted climate change. Dealing with climate change won't come out of their purse. The banking industry also did a good job of leaving all the booze bottles lying around after the party for someone else to clean up.
I suspect the die is cast. If climate is changing, we're just going to have to deal with it. It's not clear to me that we'll be better off swerving wildly and rolling the car over before we go over the cliff, than just accepting reality and going over the cliff upright.
I'm extremely skeptical that the alarmist agenda is good for anyone not partaking in the alarmist spoils. The kind of radical change required to deflect this outcome usually takes decades or generations, if you're trying to preserve the effectiveness of what already works.
Throwing away what works too lightly could easily be worse than future challenge of coping with climate change.
Banks are inherently unstable, they run at 20-50 times leverage. Small losses can wipe them out. They can only do this because they are backstopped by the US Government to the tune of trillions.
It's true, and this is the place to start thinking about a complex world. Why do you stop there? The current loop between your amygdala and your cortex creates a magnetic field that shuts down everything in between? What a peculiar failure mode.
In game theory, a dominated outcome never transpires, so you proceed as if that eventuality is entirely precluded in the game tree. With prudent leverage, there's no problem here. When prudent leverage shades into reckless leverage, there is a problem here.
You get the same problems with the MAD doctrine from the 1960s. Nuclear exchange is a dominated outcome until one side or the other succumbs to despotic insanity.
Strange mathematical levitation is also observed in the chicken and egg problem.
Therefore what exactly? Change involves risk, and a society that can't organize itself to embrace prudent risk leaves a lot more on the table than the one which risks a few stubbed toes. I don't see how we'd be busy fabricating chips at 20nm design rules on the back of a depository banking system. Perhaps over a thousand years, we could evolve the social capital to make this work. Any takers? We've been evolving social capital in our current system since the invention of coinage. East Germany stumbled for twenty years when abruptly brought in from the cold, despite lavish support from a rich uncle. But we all know the Germans are lax in confronting a challenge.
The boundary between prudent risk and reckless risk needs to be policed vigorously so that it doesn't degenerate into a Taleb tail consisting of terminal events with non-zero probability.
This is a hard problem. The core of the problem seems to be mobilizing political sentiment to do this (many stones remain unturned) when political will in society is dominated by whichever group is presently experiencing the most outrageous short-term profit.
The banker's prayer: Lord make me prudent, but not yet!
I halfway agree with Taleb. Perhaps we can keep the bonuses, but have the creditors award them rather than management. A sufficiently large bonus ought to imply that the bankers are so brilliant that nothing could possibly go wrong. In awarding a bonus at this threshold, the creditors would be required to sign documents waiving their bail-out privileges (which should directly impact their reserve requirements for their own operations). The waiver could be on a sliding scale. I'm not sure how to optimally engineer the transfer function. At what point does a bonus imply competence? $10 million? $100 million? Hard to say. We could get tangled up in argument. Perhaps best to leave the system alone. It seems to be working so far.
I second the motion, but with half the pixels. That's why I'm crawling this thread in detail, because I'm looking for a way out of 10.04. I'm sure it will end up being OpenCL that's the FreeBSD deal breaker.
It's the classic trap of optimizing the solution domain instead of the problem domain. The center of the usability world is not the computer, but the human skills and limitations of the person interacting with the computer.
A computer ought to be like a good bartender, who just knows on the first look that you aren't going to start with a parasol sticking out of a maraschino cherry. It ought to greet me with an offer to try out a unique Slivovitz or the vintage Calvados rather than offering me the pitcher of Coors Lite for one buck less than yesterday as if I care.
There's a guy on Wikipedia with nearly 800,000 edits. Shouldn't the computer make certain assumptions about his work process rather than popping up an interface suitable to his grandmother? If I sit down at a computer I've never used before and plug in my iPod, shouldn't it notice that I've never listened to a three minute pop song since I bought the device, but I do have 16GB of hour long lectures in the areas of technology, psychology, politics, history, and economics? Should my 30 years of keyboard experience not be taken into account? Or my 100-500 Google searches per day, 300 days per year, for the most of the last decade?
The bartender should just know that I need tabs and desktops, or failing that, some reasonable way to spread out.
The ultimate human assistant is the one who knows exactly how much bandwidth you have available, and when and how to interrupt you with new information or a better approach.
The interface I deserve is the one designed for the F35 fighter pilots where they actually do give a shit about your cognitive limits and making it possible to reach them. The start menu is just another deck chair on a biplane. I'm sick of interfacing with the computer. Wake me up when the computer interfaces with me.
Don't confuse Montreal with Quebec. And don't count on smiles and beneficence in the far east end: that's where the Quebecoise live who wish Montreal was more like the rest of the province.
Here's my idea: figure out who the politicians are in the neighbourhoods where these people keep their homes and families, then organize to get those politicians kicked out of office at the next election. Make these seats incumbent dead zones until the laws are strong enough to deter this kind of behaviour without calling in the space drones.
Slashdot Mirror
True, but why did you launch directly into automata? A hash-based bloom filter k=1 which tests each word against a bit vector of 8KB (64kbit) in size (L1 friendly) will flag roughly one false positive out of every 40 words (p=1600/2^16) for further testing. On the first true positive, the algorithm terminates, so this does not contribute to asymptotic run time. Depending on your CPU, the bloom filter might cost up to twenty instructions per character processed including amortized detection of word boundaries.
Then you need to include additional processing triggered by 2% of words evaluated which can also be done with efficient hash methods (likely using larger non-L1 data structures). If you haven't banned any 1 or 2 character words, you can skip these completely. The second phase processing won't add up to much.
Note that you should use a randomized hash function so that there is no predictable worst-case message available to miffed DOSers if you think this is a task worth doing well. This costs you at initialization time (dynamic instead of static bitvector), but is free thereafter.
The easiest dodge for the person writing the message is to replace every space with an infrequent character such as 'z' if the words are hard to read written with no spaces at all. Or you can go full camel case. If the rule for what determines a word boundary becomes shifty, you're into full-blown spam filter hell. Of course, in such a country, shifty word boundaries could cost you your left one, so parse prudishly if you value your thumbs.
In hindsight, perhaps, this is all clear. At the time, would you have bet your house on the proposition of 386BSD remaining unscathed if the BSDi lawsuit had come to a different outcome? But wait, I have a reference.
From Open Sources: Voices from the Open Source Revolution:
Yeah, totally clear how 386BSD was free and clear of the legal fog of war. And a huge debt owed by everyone to Marshall Kirk McKusick and friends who fought this battle on our behalf while Linux thrived under the legal radar.
In my own view, Linux had a crazy-making anthill culture, which appealed to many young coders with more energy than brains. But you know, I wouldn't bet against energy in retrospect. The annual ipchains rewrite boggled my mind. Not my cup of tea. An even crazier splinter group made hay with PHP, breaking just about every rule of thoughtfulness and elegance known to God and man. And look where that got them: pretty damn far.
I would personally, however, have jumped on the BSD wagon at the time had it been able to promote a coherent vision of life after lawsuit. What would be the balance be now if BSD had gathered twice as many elitist greybeards into the fold? I have a feeling it would have continued to lag in the department of crappy consumer product device drivers, compromising a major defection path from Windows 98. Greybeards don't do popularity worth a damn.
Debian zealots notwithstanding, Linux quickly became popular enough to become a willing host for binary blobs.
He's right in one sense: the first thing a motile organism achieves is the ability to swim in the direction away from its waste stream.
With seven billion people on the planet, and rapidly rising standards of living in China and other parts of Asia, with coal increasingly augmenting the petroleum plateau, I'd be shocked if every year didn't set a new record for CO2 emissions.
What would be unprecedented here is a global consensus among rich and poor, east and west in preventing business as usual from raising the bar on CO2 emissions every year for the next half century--which is exactly what would have happened had the earth's CO2 levels not been precisely balanced at a precarious tipping point as science presently tells the story. If we pull this off, we'll be manufacturing a political consensus out of whole cloth such as never before witnessed on this blue marble.
On another note, I don't get this beautification of scientific consensus as the second coming of fast food culture: science is, and always was, a slow food movement. It takes decades or centuries to reach secure conclusions concerning systems as complex as the earth's climate. I think this is a lot like a doctor who discovers a new disease model, then immediately proposes an extremely radical treatment of unknown severity and consequence.
Via Wikipedia:
Amputated at the thigh IIRC by another doctor who shows and takes responsibility. In the long run, these interventions become routine, and the consequences become understood and mitigated.
Is there any evidence that we can fundamentally shift the global economy away from fossil fuels on a radical program without incurring large and unknowable risks to geopolitical stability in doing so?
The paint is still wet on climate science. Be careful what you wish for. And don't write me off as a club foot surgery denier. The old day-glo Wired was my personal hot tub: I'm a card-carrying techno-optimist. Politically, however, I'm extremely wary about any combination of alacrity with wet paint. Apollo 13 was pretty much the historical high water mark on smooth sailing amidst a crash program instigated by handshakes among our political overlords.
In every experiment I've read about, the entanglement is known by how the particles are created. But here's the question I never see answered: is it possible, given two particles you know nothing about, to prove the particles are not entangled?
If you can't prove any given pair of particles are not entangled, then perhaps entanglement is the natural state, and particles known to not be entangled (by some exogenous information about their creation or history) would be the exceptional case.
In cryptography, the entanglement of key bits and cipher bits quickly becomes so diffuse you can pretend it's not there, for almost all practical purposes.
If all particles became entangled at the time of the big bang, then maybe there really isn't such a thing as unentangled particles. Now, I don't know nearly enough about physics to know how this suggestion could be easily blown out of the water, but I know enough about philosophy to puzzle about why this very basic aspect of the paradox consistently escapes clear explanation.
Perhaps the only entanglement one can observe is the one where you know something ahead of time. Then to say that two particles are not entangled is more a statement about your baseline ignorance than it is about the particles.
I suppose what you are saying is that people ALWAYS exceed the speed limit unless the speed limit exceeds the reigning land speed record, but that such a speed limit could never be adopted by any social process even though Germany has in fact adopted something not entirely different.
When the sun finally goes red giant, I'm not entirely sure the damage to the planet from nuclear energy will actually be worse than estimated.
I'm pretty much 100% certain that somewhere between 1950 and 2100--as things are presently progressing--we'll gain enough engineering and political competence to make nuclear energy a safe alternative relative to any sane norm in these matters, if by then we still wish to pursue it, which is highly doubtful, but not impossible.
Evidence now appears overwhelming that 1970 fell short of the mark and that 2010 has yet to complete its homework assignment.
Note: you can actually see an impressive long march towards human political competence through a thousand year aperture.
What is "this" exactly? The fact that they now hold the patent, and might someday do something with it? Actual power obtained from this patent after surviving a legal challenge by a motivated and cash rich adversary?
This kind of circular reasoning is what tips the game theoretic balance against having the patent system in the first place.
I was pleased with the discussion about this kind of dynamic in Satz on Markets.
The old view that transactions take place between voluntary participants in the absence of negative externality is pretty much dead and buried given the complexity and interconnectedness of the modern economy.
That's what Microsoft has become: the office work flow utility. Microsoft worked very hard to become such a utility, all the while distracting the audience with the other hand lip-synching the word "innovation" which as everyone knows is mainly MIA.
They will never describe themselves this way directly in any media that gossips to the DOJ. You just have to look at their stock performance and the people who invest. It's obvious.
If it looks like a utility through the moneyscope, it is a utility.
If you ask a group of people to self-assess for leadership skill, eighty percent report above average leadership skills. What seems to be happening is that each person defines leadership as heaps of whichever skill component he/she happens to possess, and not so much of the skill components he/she lacks. They are all telling the truth with respect to variable criteria.
So I'm wondering, does autism define "comprehend" the same way a non-autistic person does? I read about half that speed, and I don't feel limited by word recognition, but more by the multiple processes of figuring out where the author is coming from (or not, if the agenda is to apply lipstick to a mental vacuum). There are so many layers to discourse analysis it's hard to list them all.
"Comprehend" could mean retaining information points presented as fact. Or it could mean assigning the dribble of factoids into mental categories "pulled from ass", "brandishing urban legend", "regurgitated from recent popular news story", "manufactured in a pique of convenience", "seduced by right-thinking glean", "outright deception", etc. It's a lot of work when reading to man the airport scanner of psychological bogosity.
Furthermore, these assessments are fluid and tentative and require a large working reserve of rewritable storage. A model of the author as a reliable or unreliable human being is formed, if the assessment is to care enough to do the work.
This last effect is most obvious watching movies. I give weak passing grades to diverting films I couldn't care less about. If the movie gets just enough better that I start to care about what it might have been, that's when my harshest judgments are unleashed: I've entered into the punative "made me look" valley where I actually turn on the critical machinery--often to discover not entirely quickly enough that it was a false start. The subset of the movies that make me care and then reward the bother is where I start giving out decent scores.
Sometime when reading I turn the page, and go "ugh" inside and then feel the overwhelming urge to skip forward half a page or a whole page, or both pages. Then I go "how can you _know_ all this text is worthless in a tiny fraction of a second after the page flip?" So I go back and slog through it and sure enough, in the vast majority of cases, my instant assessment was right on the money.
Many of the long-winded essays linked from aldaily.com are particularly challenging in this regard. Some of those writers are talented enough to go on for page after page saying hardly anything at all, while defeating the immediate "this is vacuous crap" quick page-turn self defense. It appears that there is a high art to saying nothing in such an elaborate and convoluted way that busting the vacuity of the prose reduces me to my real-time reading speed.
I once read a piece, Kirkegaard I think, about chasing a bug around a desk with a pin while enduring immense boredom in the classroom. The humanities is where you learn to wield the pin, and make your reader perform as the bug. Not always, but fairly often. What to these people does the word "comprehend" actually mean?
That's precisely what they want to know. Giving a different answer on every web form lowers your learned-helplessness score to almost zero. They won't want your business. You might not blindly pay charges you don't understand for services you didn't know you needed.
These bonuses are the banking equivalent of free drinks at Vegas. No one drinks free who isn't losing far more than the drinks are worth.
You think the smart people will simply head off to the smart banks that don't go in for this nonsense? Then how is it that the credit card companies have convinced retailers not to discount cash? People who pay cash miss out on all the great CC swag.
Ultimately they get you coming or going.
"Geek IQ" is a polite synonym for mastering alternate life pursuits not entirely like living. However, such a test ought to know the difference between nerd IQ and geek IQ.
The question about "Heroes" is the moral equivalent of the birthstone questions in the science category of the original Trivial Pursuit.
Every category had some bird food. In Sports, you just keep picking Bath Ruth or Mohamed Ali whenever they come up, you'll get one eventually. For a science ignoramus, the closest surrogate to low-hanging fruit is substituting astrology for astronomy.
Google is my friend, too, yet I would have been ever so grateful for the tiniest social grace of sparing me yet another Google result set.
I'm pretty sure Lewis Thomas in Lives of a Cell (or possibly Et Cetera, Et Cetera: Notes of a Word-Watcher) comments about the sad decline of the elder statesmen: he hasn't forgotten anything so much as piled it so deep in the attic he can't find it without a substantial jog.
For about five minutes a year ago I knew what Diaspora was. Then it went directly to the Lewis Thomas attic of things I can only possibly remember once reminded.
Hard to understand, I guess, when you're twenty two.
The world has become a kind and gentle place. Used to be the best advice was to sever your penis. Have you ever thought of signing up for Root Strikers? You're a natural. No, wait, I have it backwards: your advice is a gilded path to office in Washington.
Outside of this post, there sure seems to be a lot of people on this thread suffering from reality shock.
If assassins had a union, the union would terminate any member who chills business by snuffing out any prospective customer who shows up, unless everyone is pretty sure you kept your solicitation close to the vest.
I suppose it's like the priest who hits a hole in one on the golf course when he should have been preaching: Who can he tell? When you run with a crowd where you can reveal your murderous solicitations as insurance against the instant double-cross, we call it organized crime.
In any case, with this kind of intimidation, there's often a tipping point where there's too much to suppress and too much risk in doing so. When the heads pile up to the sky, it becomes a political issue.
On a podcast the other night, I heard that stern discipline of children promotes frequency and skill at lying. Some parents care enough to give their children a big headstart on a lucrative career path.
The vast majority of blue screens I had with 2000 were hardware related: bad chip sets (Via KT7A), bad memory (many), bad capacitors (anything from Asia), bad disk drives (Fujitsu, IBM). Another class of blue screens pertained to immature device drivers. Video cards in particular were a fast moving niche, and developing for Windows 2000 was not yet a common expertise. I used 2000 for a long time because I never upgraded to XP. There was no significant stability problem.
Hardware maturity has also come a long ways since 2000. As a piece of software, I never felt I had stability problems with Windows 2000. You were somewhat at the mercy of the device drivers installed. Fail-safe sandboxing comes with a performance penalty that was less attractive on a 700MHz Pentium III with 512MB of DRAM. I think an order of magnitude estimate on instability of 2000 vs Windows 7 is gross inflation. If you can find a mainboard from 2003 that hasn't leaked its mojo, install Windows 7 and see how its stability rates.
I did have a fair number of blue screens under NT that were entirely NT's fault. The NT AppleTalk networking component was a major puntfest on my home network.
What keeps me on Windows: any single killer application I can't replicate under Linux.
What keeps me off Windows: imprisonment of my work product in proprietary vendor data formats.
One is a short term liability, the other is a long term liability. Migration continues. Speed of the migration would be enhanced if all geeks unite to eschew the nVidia binary blob.
The computer capable of that level of introspection and inference would snort at your silly fashion bias toward functional languages. The main calling card of functional languages is to offset weakness in human cognition. The human brain struggles to convert a functional specification into an optimal state machine without dropping a stitch. Kasparov and others complain about computer chess precisely because a well-tested adversary never drops a stitch, or so rarely that chess programmers have a dozens of other things to worry about first.
You do realize that the primary virtue of a functional language is purity in the specification domain and that it offers no fundamental advance in the execution domain?
There are two halves to the specification domain: algorithmic correctness, and shaping the performance/resource envelope. Prolog is a fairly reasonable specification language for algorithmic correctness, but almost completely useless at shaping the performance/resource envelope. Ever seen a smart phone OS programmed in Prolog?
There needs to be a word for this particular cognitive bias. This is the cognitive bias that if there's enough food on the planet, no one should starve, neglecting only the distribution challenge modulo politics, history, culture, and human nature.
In a world dominated by programming languages optimized for algorithmic correctness, all our problems will miraculously go away, because all those potent algorithms will sort out the performance/resource envelope without further input of blood, sweat, and tears. Nice vision.
That day will arrive when I specify the desired solution as a shortest path and the computer responds, "no can do, but would you settle for nearly as good almost all of the time under modest stochasticity assumptions in the underlying graph in near real-time to the largest feasible problem size as practically bounded by performance bounds elsewhere in the application feature set as they presently exist for the targeted user base?"
And I will go, could you break that down into smaller pieces? I'm out of practice thinking that hard.
On the transistor topic, it's kind of stupid to neglect the power distribution tree. Idle execution units don't leak if the master valve is slammed shut. In future we can have a much larger set of execution units optimized for different tasks, and only use the one that's needed for a heavy lifting loop.
You're already seeing the shift to dark silicon with the introduction of the ARM A5 as a companion dog to a bigger OOO furnace. One or the other CPU is shut off completely at any given time. Hard to leak power that never arrives.
Those of us that have an open mind don't put such a low price on our precious gift. From what I've read, the ratio of show stoppers to features that might possibly make my life better (after a significant accommodation period) is about ten to one. My number one productivity feature right now is having nine desktops on two large displays. I've not read a single Unity review that says "expect a great experience with the dual-head multidesk".
Open mindedness is not carte blanche to compromise what already works to pursue a pot of gold at the other end of the usability rainbow. To make matters worse, I can't even see this particular rainbow, much less the pot of gold.
Ubuntu screwed this up. There should have been at least a full two year period where a feature complete Unity existed in parallel with the tried-and-true so that the open minded could pick their spot for making the switch.
If you've only got one ball in the air, and you don't mind this ball being adaptation to Unity, then I question whether you're an open minded person at heart, or just some chump who thrives on the reckless pursuit of novelty.
Also apropos, check out Ch.4: The faults of others by Jonathan Haidt on myopic self-interest before yanking morality play red lever.
Zinga will still attract talent, but the contracts will have thicker prose about termination and vesting conditions, so it will be much harder to pull this stunt a second time. But perhaps at this stage of growth they are beyond that.
I watched Politics, Strategy, and Game Theory last night, which talks about Grim Trigger and the conditions under which, in iterated prisoner's dilemma, you care more about the future than defecting in the present moment.
It's a competent lecture with no great pizzazz.
Here's a fairly nice piece by an undergraduate I stumbled upon brushing up on Grim Trigger: Debunking the Prisoner(slashcode fuckup)s Dilemma on Robert Axelrod(slashcode fuckup)s Emergence of Cooperation among Egoists and why cooperation is a lot more common than the shallow analysis would have you believe.
I really wonder what payoff matrix he constructed to author that blog under no fixed identity. I found a Tweet referencing the site by the apparent author with one or two clues about his circumstance.
The punishment for Zynga in future iterations are employment contracts with a lot less room to wiggle if they screw up future hires. The reward in the present iteration is yanking back a substantial chunk of the entire company.
If the quiet vestors really aren't showing up and pulling their weight, it doesn't seem great to let them get away with that either. I don't think Zynga's presumption is that they can't fight this, but more like "the effort involved will be a shock to their lazy asses" so they are likely to settle without going ten rounds.
Yeah, after Y2K there were headlines screaming about the failure of failure wondering if the preparation had been worth it after all.
I'm not sure assigning blame is the first order of business. I would slice things up a little differently. One of the main reasons that China's output is growing so rapidly is that the developed world has shifted manufacturing offshore. Much of the CO2 released from China is making the consumer goods we consume in America and Europe.
By the same token, the high levels of emissions in America and Europe correspond to the majority of technological progress over the last century, which the whole world shares. The low power cell phones in Africa were made possible by an immense technological investment elsewhere.
Which body organ do you blame? The more interesting analysis is whether we're using energy wisely. America especially uses a lot of energy rather stupidly. We build sprawling cities with massive road systems for single occupant vehicles instead of denser cities with excellent public transit. Some of those decisions might have panned out differently if gasoline prices had been a lot higher a lot sooner. Detroit didn't want it that way. Detroit wanted climate change. Dealing with climate change won't come out of their purse. The banking industry also did a good job of leaving all the booze bottles lying around after the party for someone else to clean up.
I suspect the die is cast. If climate is changing, we're just going to have to deal with it. It's not clear to me that we'll be better off swerving wildly and rolling the car over before we go over the cliff, than just accepting reality and going over the cliff upright.
I'm extremely skeptical that the alarmist agenda is good for anyone not partaking in the alarmist spoils. The kind of radical change required to deflect this outcome usually takes decades or generations, if you're trying to preserve the effectiveness of what already works.
Throwing away what works too lightly could easily be worse than future challenge of coping with climate change.
It's true, and this is the place to start thinking about a complex world. Why do you stop there? The current loop between your amygdala and your cortex creates a magnetic field that shuts down everything in between? What a peculiar failure mode.
In game theory, a dominated outcome never transpires, so you proceed as if that eventuality is entirely precluded in the game tree. With prudent leverage, there's no problem here. When prudent leverage shades into reckless leverage, there is a problem here.
You get the same problems with the MAD doctrine from the 1960s. Nuclear exchange is a dominated outcome until one side or the other succumbs to despotic insanity.
Strange mathematical levitation is also observed in the chicken and egg problem.
Therefore what exactly? Change involves risk, and a society that can't organize itself to embrace prudent risk leaves a lot more on the table than the one which risks a few stubbed toes. I don't see how we'd be busy fabricating chips at 20nm design rules on the back of a depository banking system. Perhaps over a thousand years, we could evolve the social capital to make this work. Any takers? We've been evolving social capital in our current system since the invention of coinage. East Germany stumbled for twenty years when abruptly brought in from the cold, despite lavish support from a rich uncle. But we all know the Germans are lax in confronting a challenge.
The boundary between prudent risk and reckless risk needs to be policed vigorously so that it doesn't degenerate into a Taleb tail consisting of terminal events with non-zero probability.
This is a hard problem. The core of the problem seems to be mobilizing political sentiment to do this (many stones remain unturned) when political will in society is dominated by whichever group is presently experiencing the most outrageous short-term profit.
The banker's prayer: Lord make me prudent, but not yet!
I halfway agree with Taleb. Perhaps we can keep the bonuses, but have the creditors award them rather than management. A sufficiently large bonus ought to imply that the bankers are so brilliant that nothing could possibly go wrong. In awarding a bonus at this threshold, the creditors would be required to sign documents waiving their bail-out privileges (which should directly impact their reserve requirements for their own operations). The waiver could be on a sliding scale. I'm not sure how to optimally engineer the transfer function. At what point does a bonus imply competence? $10 million? $100 million? Hard to say. We could get tangled up in argument. Perhaps best to leave the system alone. It seems to be working so far.
I second the motion, but with half the pixels. That's why I'm crawling this thread in detail, because I'm looking for a way out of 10.04. I'm sure it will end up being OpenCL that's the FreeBSD deal breaker.
It's the classic trap of optimizing the solution domain instead of the problem domain. The center of the usability world is not the computer, but the human skills and limitations of the person interacting with the computer.
A computer ought to be like a good bartender, who just knows on the first look that you aren't going to start with a parasol sticking out of a maraschino cherry. It ought to greet me with an offer to try out a unique Slivovitz or the vintage Calvados rather than offering me the pitcher of Coors Lite for one buck less than yesterday as if I care.
There's a guy on Wikipedia with nearly 800,000 edits. Shouldn't the computer make certain assumptions about his work process rather than popping up an interface suitable to his grandmother? If I sit down at a computer I've never used before and plug in my iPod, shouldn't it notice that I've never listened to a three minute pop song since I bought the device, but I do have 16GB of hour long lectures in the areas of technology, psychology, politics, history, and economics? Should my 30 years of keyboard experience not be taken into account? Or my 100-500 Google searches per day, 300 days per year, for the most of the last decade?
The bartender should just know that I need tabs and desktops, or failing that, some reasonable way to spread out.
The ultimate human assistant is the one who knows exactly how much bandwidth you have available, and when and how to interrupt you with new information or a better approach.
The interface I deserve is the one designed for the F35 fighter pilots where they actually do give a shit about your cognitive limits and making it possible to reach them. The start menu is just another deck chair on a biplane. I'm sick of interfacing with the computer. Wake me up when the computer interfaces with me.
Don't confuse Montreal with Quebec. And don't count on smiles and beneficence in the far east end: that's where the Quebecoise live who wish Montreal was more like the rest of the province.
Here's my idea: figure out who the politicians are in the neighbourhoods where these people keep their homes and families, then organize to get those politicians kicked out of office at the next election. Make these seats incumbent dead zones until the laws are strong enough to deter this kind of behaviour without calling in the space drones.
You have to apply pressure where the power lives.