Today, both are acceptable and both are found in dictionaries. There is no appreciable difference in meaning. Whether you are a cardsharp or a card shark, you are still a shifty character.
No, as I said "card shark" sometimes means a good player and only sometimes has shify connotations (as opposed to cardsharp).
See, e.g, Webster's: shark: 1 : a rapacious crafty person who preys upon others through usury, extortion, or trickery 2 : one who excels greatly especially in a particular field
I predict that bzip -9 corpus.txt will be the winner of your contest, or more accurately, followed by strip bunzip2 && bunzip2 corpus.txt.bz2, where the pair of (bunzip2 corpus.txt.bz2) is the program.
I'd think lzma would beat it (along with other general-purpose compressions algorithms that have better compression rations than bzip2).
But I wouldn't be too shocked if a superior special-purpose algorithm could be written that worked better on (say) english-language text but far worse on everything else. And such a thing is likely to be somewhat interesting.
Though I still think any AI that acts (to the average human observer) deterministic, let alone computable, is doomed.
Maybe it worked, maybe it didn't. When was the last time you saw a can of Burma Shave on the store shelf?:)
They've been bought by American Safety Razor, but the brand is still around (almost entirely because of these ads). They even ran some of the old-style road signs in North Carolina about 5-6 years ago.
Amusing that the brand is now attached to brush shave-cream, since Burma Shave was one of the original brushless creams and often made fun of the brush ("Shaving Brushes/You'll soon see 'em/on a shelf/in some museum/Burma Shave")
Most of the ads would have 4-5 signs, then the "Burma Shave" tag sign at the end; e.g. "Dinah doesn't/Treat him right/but if he shaved/Dinah might/Burma Shave".
But there was one series that omitted the Tag, showing how ubiquitous these signs once were: If you don't know who we are you haven't travelled very far.
The original signs ran from the 1920s-1960s.
And in the mid-80s someone put up a bunch of sets that said: Farewell O verse Along the road How sad to see You're out of mode.
but as I said, the late 1990s saw the return of some Burma-Shave signs.
Sure, as long as that person gets shot, and they accidentally bring him to this research facility instead of a hospital...
The U Pitt medical center (which the Safar Institute is part of) is a lot like the Johns Hopkins Medical Center or other research medical centers. It's a research hospital, but because it has many of the top doctors in the area it also gets severe cases (including massive trauma cases) that require special skills.
Also, like all hospitals, there's an element of proximity; if you get shot in the Oakland neighborhood of Pittsburgh (where the University is), they're probably going to treat you in UPMC and not rush you bleeding to some other hospital farther away.
So it's pretty likely that a mortally wounded gunshot victim would wind up there.
The Safar Center was doing these experiments successfully in 1996.
I have no idea if they've recently done yet another incrementally longer period of exsanguination, as the article doesn't mention the time or a journal article name or anything.
Bottom line, you need incremental backups for data reliability. Doesn't matter how you do it, you can do it on top of RAID 5 to give you more peace of mind if you want, but it's not really necessary. Instead, at a bare minimum, you must be able to go back to several points in time to recover as recent of data as possible.
This document describes a method for generating automatic rotating "snapshot"-style backups on a Unix-based system, with specific examples drawn from the author's GNU/Linux experience. Snapshot backups are a feature of some high-end industrial file servers; they create the illusion of multiple, full backups per day without the space or processing overhead. All of the snapshots are read-only, and are accessible directly by users as special system directories. It is often possible to store several hours, days, and even weeks' worth of snapshots with slightly more than 2x storage. This method, while not as space-efficient as some of the proprietary technologies (which, using special copy-on-write filesystems, can operate on slightly more than 1x storage), makes use of only standard file utilities and the common rsync program, which is installed by default on most Linux distributions. Properly configured, the method can also protect against hard disk failure, root compromises, or even back up a network of heterogeneous desktops automatically.
THINK about what you are saying. Youre telling this guy to piss off his boss to the point where he may be fired or considered not a team player and passed over for opportunities, and then sue them. Ok so the guy gets fired, with what money is he going to sue them?
I agree, suing them is unrealistic. It's potentially worth documenting anyway, if they're going to fire you you can point out that it's illegal and essentially settle with them--but at that point relations are going to be SOUR, so it's probably only a stopgap to keep you working while you job hunt.
I've heard the retort "well if the company does things like this, this isnt a company you want to work for." It just does not hold water. The company might be awesome in all other respects, but they don't feel that software piracy is something to be concerned about.
This is BS. It's one thing if a company has some policies you don't want, asks you to work more weekends than you'd like, etc. I could maybe see overlooking them installing illegal copies of software for you to use--I'd have a problem with that, but I haven't been in that situation to know if I'd actually walk or not.
But it's an entirely different thing if they order you to commit illegal actions. It's not worth it, especially since you know it's illegal--you're accepting personal liability beyond the payment you're getting from the job.
However, I would also agree that a jury would probably feel the same as you in most circumstances and acquit (but this would be from a human standpoint, not from a point of law).
No, it's a point of law. Jury nullification is legal and well-established through court rulings and precedence. It's not codified, but neither are a lot of other court precedents.
You're on the right track, but the problem with your analogy is bar mitzvahs don't product children. (OK, so the really good ones do).
That's the real point of marriage - to establish care and lineage for children. Everything else is window dressing, tossed in for legislative convenience.
You're wrong. The purpose of marriage is to make a lifelong commitment to another person. Read the freaking vows some time.
It used to be associated with children, but these days (heterosexual) marriages without children are not _that_ rare--about 1 in 5 marriages are childless, including 10% of marriages where the wife is over 40 years old. Children outside of marriage are, obviously, quite common.
The legislation for children needs to protect those outside marriage. And there's no good reason to apply it to people who wish to marry but not have kids.
A national argument about gay adoption makes sense. An argument about gay marriage is just silly. Marriage should be a private contract between two individuals, and the state shouldn't be involved.
Also, I'm a little skeptical of any report that claims to be able to tell the difference between a consumer demanding reports skewed to his beliefs, and a consumer demanding objective coverage. That seems loaded from the start.
You'd have to read the report to see exactly how they arrive at this conclusion, but a large part of it is based on (a) seeing established publishers/outlets enter new markets and observing the changes in their coverage and (b) observing the same publication's relative success in areas with different biases.
For (a), e.g., Murdoch's Boston publications are far more liberal than his average publication, and liberalized rapidly after entering the market.
For (b), you can take a look at any baseline publication with a national market and see that its sales figures correlate with the local bias--so, e.g., NYT sells better in Austin than Houston, and better in San Francisco than Austin, and vice-versa for conservative publications. It's tough to argue that NYT is somehow more objective in San Francisco than in Houston.
NYT is actually closer to the average than Fox News is
Consumers select media that aligns with their beliefs.
My logic would say that if those two statements are true, FNC would be marginalized and shrinking, and the NYT would be prominent and growing.
Umm, newspapers are almost all shrinking rapidly. Cable TV is almost all gaining viewers. Those factors dominate the relatively slow changes in liberal/conservative bias in the general public.
Are you saying that there are still more viewers of liberal-biased media than conservative-bias media if you take them as a total?
I'm saying that the average bias of the media is farther to the left of Fox than it is to the right of the NYT. I'm not sure it's fair to call it liberal-biased if it is in agreement with the general population.
But yeah, if you add up CBS + NBC + ABC + CNN + PBS + internet + print +... then even a growing Fox doesn't skew the overall numbers that much, and it's more of an outlier than NYT is (though neither are extremely close to the average).
the American people are by and large more liberal than the members of Congress
Which decade are you talking about? It can't be the current one. More Americans have identified themselves as conservatives than Liberals for a very long time.
But that stat has nothing to do with the quote that you're trying to refute.
About the only place that the New York Times possibly reflects the public's beliefs is in NY City itself and maybe SF, DC or LA
Yes, this is another point made in the article I cited; not only does the media as a whole trail public trends in liberal/conservative bias, but within regions it also tracks closely (e.g. a liberal-owned paper in Texas tends to be much more conservative than the average liberal-owned paper, just as a conservative-owned paper in Massachucetts tends to be more liberal than the average conservative-owned paper). Indeed, if you aggregate the media outlets in a particular region, their total bias aligns very closely with the region's bias. This turns out to be an excellent way to test your bias measure.
Note that I didn't say that NYT isn't left of the average opinion, merely that it is closer to the true average than Fox News, contrary to the assertion of the parent article that NYT is twice as far from the mean bias as Fox. Neither is particularly close to the average, though.
How can you say that viewers follow the shows that align with their beliefs, yet say that NYT is closer to most Americans than FNC? FNC is wildly successful, so if viewers demand coverage skewed to their perspective, Americans must be more like FNC than the liberal outlets
You're trying to mix up two different points here.
1. Consumers select media that aligns with their beliefs. This results in media bias basically in line with their beliefs. If you look at the total media bias (yes, including NYT and Fox News, but also CBS, NBC, ABC, CNN, all the newspapers, etc) then it aligns very closely with the observed bias of the general population.
The paper actually goes further than that, showing that the bias in regional markets also aligns with the consumer bias in those markets--so a Murdoch-owned paper in, say, Massachucetts is much more liberal than a Murdoch-owned paper in Texas, and vice-versa).
2. Because of (1), the skew that the parent post alleged is in reality backwards; the parent argued that Fox is half as biased as NYT, but that's simply not true. The actual average American is more liberal than the average congressperson, and it turns out that far from being twice as far off, the NYT is actually closer to the average than Fox News is.
Of course, this study has deep methodological problems that have been discussed to death in the statistical reporting community (See, e.g., Measuring Media Bias, Michael Cardwell, George Mason University, March 1, 2005). The consensus seems to be that the studied media outlets trail their consumer's tendencies in this area--that is, consumers do not, in fact, demand objective coverage, but rather demand coverage skewed to match their views, and media outlets tailor their product to consumer demands. And that changes in consumer bias precede (and drive) changes in media bias.
One of the major findings is that the American people are by and large more liberal than the members of Congress (in large part because conservatives tend to vote more than liberals, possibly because of age correlations), so comparisons to members of Congress don't tell you whether the media is skewed relative to the general population--and, in fact, it appears that it is not.
The second upshot is that, since the general population's conservative/liberal leanings are farther to the left than a study of members of Congress would show, it turns out that not only is the media on the whole in line with the public's stance, but that the New York Times is far closer in line with the public's beliefs than is Fox News.
Note that none of this is meant as a vindication of any journalistic integrity or objectivity; on the contrary, it seems to be basically a result of the media outlets following the dollar and trying to present the news as people want to hear it.
I never said anything about causation. I was simply refuting the parent's argument that (A) "Jobs out of control" implies (B) "Apple makes a ton of money".
From (A) and (~B), one can infer that (A->B) is false.
It's just not that simple. The right front management might do better, but we have no proof of that. I suspect it's true, but on the other hand the wrong management team could do far worse.
For a basic single-floor house without basement and pre-laid foundation...
If all of my programming tasks were simple designs, lacked basic features, and had the hardest part of the work done ahead of time, I'd be a lot faster too.
Frankly, you have no clue what you are talking about nor to whom. I don't "invest", I'm a professional trader - mostly US treasury future triangular cash/future arbitrage
Yes, like I said, you try to make money timing small market fluctuations.
When the final print didn't come, a large number of professional traders who get out of all or most of their positions by the end of the day got stuck with tied up capital, interest costs, and overnight risk that needed to be hedged as a result of no fault of their own.
It's not like they "got stuck" with them. They bought them, and hadn't yet sold them.
If you're making stock decisions that rely on 4-minute windows, you should expect to be screwed on a regular basis. Really, trying to time market fluctuations more closely than at least yearly terms--and more realistically, 5 year+ terms--is a crap shoot. Barring highly unusual situations, if a company was a good long-term prospect 4 minutes or 4 days ago then it's still a good prospect now. Really, how many of the positions you held made massive announcements that affected the company's prospects in those 4 minutes?
More to the point, whatever you were holding when trades stopped, you were still holding the next day, and the NYSE going down doesn't affect the value of the companies whose issues you held. Nor was your right to buy or sell shares affected, you just lost 4 minutes of service from the largest market for finding buyers/sellers.
Guys like Warren Buffett--or large fund managers-- don't speculate on minute-to-minute fluctations, and if the market was down for a couple of days it wouldn't really affect their ability to make wise investments. And tiny downtimes like this don't materially affect the fundamentals of the market system.
By nature, if you play with day-trading you're accepting a much higher risk, and when that risk materializes I don't feel that bad for you.
You can have "pretty good" software, you can have "excellent" software, you can even have "never, ever been broken" software, but you can't have "provably perfect" software in all but the most trivial of cases, (and even then things are suspect because of other potential issues: BIOS reliance, hardware manufacturing defects, etc.)
I agree with you on the hardware.
But provable software can exist beyond the trivial. See, e.g., the FOX project at Carnegie Mellon University which has implemented an HTTP server, packet filters, etc in ML, all (mathematically) proven correct.
Proving results about larger software is made much easier by doing strictly functional programming in strong statically typed languages, hence their use of ML.
They've even done research on proof-carrying code, where foreign programs come distributed along with proofs of their behavior that are verified before running them.
Now, there are some caveats: 1. You can prove that certain conditions aren't violated, or that others are always met; whether that spec meets your definition of security is up to you. 2. The FOX verification relies on the correctness of the underlying OS. There's no reason in principle that you can't write a verified OS, but it's not yet been done. (verifying a special-purpose OS may suffice for some purposes).
But the situation is nowhere near as dire as you make it out to be.
Slashdot Mirror
Today, both are acceptable and both are found in dictionaries. There is no appreciable difference in meaning. Whether you are a cardsharp or a card shark, you are still a shifty character.
0 8/06/CARD_SHARK.html - liu.html
No, as I said "card shark" sometimes means a good player and only sometimes has shify connotations (as opposed to cardsharp).
See, e.g, Webster's:
shark:
1 : a rapacious crafty person who preys upon others through usury, extortion, or trickery
2 : one who excels greatly especially in a particular field
The second definition is commonly used (as is the first), for instance:
http://www.bostonphoenix.com/archive/features/98/
http://www.news.harvard.edu/gazette/2005/06.02/03
etc
I'd think lzma would beat it (along with other general-purpose compressions algorithms that have better compression rations than bzip2).
But I wouldn't be too shocked if a superior special-purpose algorithm could be written that worked better on (say) english-language text but far worse on everything else. And such a thing is likely to be somewhat interesting.
Though I still think any AI that acts (to the average human observer) deterministic, let alone computable, is doomed.
That's backwards. A "cardsharp" (one word) is a cheater, a "card shark" is a good player, but sometimes the word has connotations of deception.
s harp
/-"shärp/
http://www.webster.com/cgi-bin/dictionary?va=card
Main Entry: cardsharper
Pronunciation: -"shär-p&r
Variant(s): or cardsharp
Function: noun
: one who habitually cheats at cards
They've been bought by American Safety Razor, but the brand is still around (almost entirely because of these ads). They even ran some of the old-style road signs in North Carolina about 5-6 years ago.
You can buy their current products at (for instance):
http://www.diamondbeauty.com/brandnames/Burma-Sha
http://store.darisimall.com/798819.html
Amusing that the brand is now attached to brush shave-cream, since Burma Shave was one of the original brushless creams and often made fun of the brush ("Shaving Brushes/You'll soon see 'em/on a shelf/in some museum/Burma Shave")
Most of the ads would have 4-5 signs, then the "Burma Shave" tag sign at the end; e.g. "Dinah doesn't/Treat him right/but if he shaved/Dinah might/Burma Shave".
But there was one series that omitted the Tag, showing how ubiquitous these signs once were:
If you don't know
who we are
you haven't travelled
very far.
The original signs ran from the 1920s-1960s.
And in the mid-80s someone put up a bunch of sets that said:
Farewell O verse
Along the road
How sad to see
You're out of mode.
but as I said, the late 1990s saw the return of some Burma-Shave signs.
Sure, as long as that person gets shot, and they accidentally bring him to this research facility instead of a hospital...
The U Pitt medical center (which the Safar Institute is part of) is a lot like the Johns Hopkins Medical Center or other research medical centers. It's a research hospital, but because it has many of the top doctors in the area it also gets severe cases (including massive trauma cases) that require special skills.
Also, like all hospitals, there's an element of proximity; if you get shot in the Oakland neighborhood of Pittsburgh (where the University is), they're probably going to treat you in UPMC and not rush you bleeding to some other hospital farther away.
So it's pretty likely that a mortally wounded gunshot victim would wind up there.
The Safar Center was doing these experiments successfully in 1996.
I have no idea if they've recently done yet another incrementally longer period of exsanguination, as the article doesn't mention the time or a journal article name or anything.
No, it isn't, and many attempted AI implementations don't make that assumption.
See, for instance:
http://www.mikerubel.org/computers/rsync_snapshot
This document describes a method for generating automatic rotating "snapshot"-style backups on a Unix-based system, with specific examples drawn from the author's GNU/Linux experience. Snapshot backups are a feature of some high-end industrial file servers; they create the illusion of multiple, full backups per day without the space or processing overhead. All of the snapshots are read-only, and are accessible directly by users as special system directories. It is often possible to store several hours, days, and even weeks' worth of snapshots with slightly more than 2x storage. This method, while not as space-efficient as some of the proprietary technologies (which, using special copy-on-write filesystems, can operate on slightly more than 1x storage), makes use of only standard file utilities and the common rsync program, which is installed by default on most Linux distributions. Properly configured, the method can also protect against hard disk failure, root compromises, or even back up a network of heterogeneous desktops automatically.
Marcus Hutter's AIXI paper provides a proof that if an agent is a good model for human behavior, and the universe is computable
You can stop right there; as far as modern science can tell, the last assumption is invalid.
THINK about what you are saying. Youre telling this guy to piss off his boss to the point where he may be fired or considered not a team player and passed over for opportunities, and then sue them. Ok so the guy gets fired, with what money is he going to sue them?
I agree, suing them is unrealistic. It's potentially worth documenting anyway, if they're going to fire you you can point out that it's illegal and essentially settle with them--but at that point relations are going to be SOUR, so it's probably only a stopgap to keep you working while you job hunt.
I've heard the retort "well if the company does things like this, this isnt a company you want to work for." It just does not hold water. The company might be awesome in all other respects, but they don't feel that software piracy is something to be concerned about.
This is BS. It's one thing if a company has some policies you don't want, asks you to work more weekends than you'd like, etc. I could maybe see overlooking them installing illegal copies of software for you to use--I'd have a problem with that, but I haven't been in that situation to know if I'd actually walk or not.
But it's an entirely different thing if they order you to commit illegal actions. It's not worth it, especially since you know it's illegal--you're accepting personal liability beyond the payment you're getting from the job.
However, I would also agree that a jury would probably feel the same as you in most circumstances and acquit (but this would be from a human standpoint, not from a point of law).
No, it's a point of law. Jury nullification is legal and well-established through court rulings and precedence. It's not codified, but neither are a lot of other court precedents.
You're on the right track, but the problem with your analogy is bar mitzvahs don't product children. (OK, so the really good ones do).
That's the real point of marriage - to establish care and lineage for children. Everything else is window dressing, tossed in for legislative convenience.
You're wrong. The purpose of marriage is to make a lifelong commitment to another person. Read the freaking vows some time.
It used to be associated with children, but these days (heterosexual) marriages without children are not _that_ rare--about 1 in 5 marriages are childless, including 10% of marriages where the wife is over 40 years old. Children outside of marriage are, obviously, quite common.
The legislation for children needs to protect those outside marriage. And there's no good reason to apply it to people who wish to marry but not have kids.
A national argument about gay adoption makes sense. An argument about gay marriage is just silly. Marriage should be a private contract between two individuals, and the state shouldn't be involved.
Also, I'm a little skeptical of any report that claims to be able to tell the difference between a consumer demanding reports skewed to his beliefs, and a consumer demanding objective coverage. That seems loaded from the start.
You'd have to read the report to see exactly how they arrive at this conclusion, but a large part of it is based on (a) seeing established publishers/outlets enter new markets and observing the changes in their coverage and (b) observing the same publication's relative success in areas with different biases.
For (a), e.g., Murdoch's Boston publications are far more liberal than his average publication, and liberalized rapidly after entering the market.
For (b), you can take a look at any baseline publication with a national market and see that its sales figures correlate with the local bias--so, e.g., NYT sells better in Austin than Houston, and better in San Francisco than Austin, and vice-versa for conservative publications. It's tough to argue that NYT is somehow more objective in San Francisco than in Houston.
Consumers select media that aligns with their beliefs.
Umm, newspapers are almost all shrinking rapidly. Cable TV is almost all gaining viewers. Those factors dominate the relatively slow changes in liberal/conservative bias in the general public.
Are you saying that there are still more viewers of liberal-biased media than conservative-bias media if you take them as a total?
I'm saying that the average bias of the media is farther to the left of Fox than it is to the right of the NYT. I'm not sure it's fair to call it liberal-biased if it is in agreement with the general population.
But yeah, if you add up CBS + NBC + ABC + CNN + PBS + internet + print +... then even a growing Fox doesn't skew the overall numbers that much, and it's more of an outlier than NYT is (though neither are extremely close to the average).
Which decade are you talking about? It can't be the current one. More Americans have identified themselves as conservatives than Liberals for a very long time.
But that stat has nothing to do with the quote that you're trying to refute.
Yes, this is another point made in the article I cited; not only does the media as a whole trail public trends in liberal/conservative bias, but within regions it also tracks closely (e.g. a liberal-owned paper in Texas tends to be much more conservative than the average liberal-owned paper, just as a conservative-owned paper in Massachucetts tends to be more liberal than the average conservative-owned paper). Indeed, if you aggregate the media outlets in a particular region, their total bias aligns very closely with the region's bias. This turns out to be an excellent way to test your bias measure.
Note that I didn't say that NYT isn't left of the average opinion, merely that it is closer to the true average than Fox News, contrary to the assertion of the parent article that NYT is twice as far from the mean bias as Fox. Neither is particularly close to the average, though.
How can you say that viewers follow the shows that align with their beliefs, yet say that NYT is closer to most Americans than FNC? FNC is wildly successful, so if viewers demand coverage skewed to their perspective, Americans must be more like FNC than the liberal outlets
You're trying to mix up two different points here.
1. Consumers select media that aligns with their beliefs. This results in media bias basically in line with their beliefs. If you look at the total media bias (yes, including NYT and Fox News, but also CBS, NBC, ABC, CNN, all the newspapers, etc) then it aligns very closely with the observed bias of the general population.
The paper actually goes further than that, showing that the bias in regional markets also aligns with the consumer bias in those markets--so a Murdoch-owned paper in, say, Massachucetts is much more liberal than a Murdoch-owned paper in Texas, and vice-versa).
2. Because of (1), the skew that the parent post alleged is in reality backwards; the parent argued that Fox is half as biased as NYT, but that's simply not true. The actual average American is more liberal than the average congressperson, and it turns out that far from being twice as far off, the NYT is actually closer to the average than Fox News is.
Of course, this study has deep methodological problems that have been discussed to death in the statistical reporting community (See, e.g., Measuring Media Bias, Michael Cardwell, George Mason University, March 1, 2005). The consensus seems to be that the studied media outlets trail their consumer's tendencies in this area--that is, consumers do not, in fact, demand objective coverage, but rather demand coverage skewed to match their views, and media outlets tailor their product to consumer demands. And that changes in consumer bias precede (and drive) changes in media bias.
One of the major findings is that the American people are by and large more liberal than the members of Congress (in large part because conservatives tend to vote more than liberals, possibly because of age correlations), so comparisons to members of Congress don't tell you whether the media is skewed relative to the general population--and, in fact, it appears that it is not.
The second upshot is that, since the general population's conservative/liberal leanings are farther to the left than a study of members of Congress would show, it turns out that not only is the media on the whole in line with the public's stance, but that the New York Times is far closer in line with the public's beliefs than is Fox News.
Note that none of this is meant as a vindication of any journalistic integrity or objectivity; on the contrary, it seems to be basically a result of the media outlets following the dollar and trying to present the news as people want to hear it.
I use EBCDIC, you insensitive clod!
Exactly how do the dead-tree readers edit a wiki?
I never said anything about causation. I was simply refuting the parent's argument that (A) "Jobs out of control" implies (B) "Apple makes a ton of money".
From (A) and (~B), one can infer that (A->B) is false.
It's just not that simple. The right front management might do better, but we have no proof of that. I suspect it's true, but on the other hand the wrong management team could do far worse.
I think if Jobs would get off his high horse and actually cede some control over his little empire, Apple would stand to earn a ton more money
He left.
They tanked.
He came back.
Profits soared.
For a basic single-floor house without basement and pre-laid foundation...
If all of my programming tasks were simple designs, lacked basic features, and had the hardest part of the work done ahead of time, I'd be a lot faster too.
Frankly, you have no clue what you are talking about nor to whom. I don't "invest", I'm a professional trader - mostly US treasury future triangular cash/future arbitrage
Yes, like I said, you try to make money timing small market fluctuations.
When the final print didn't come, a large number of professional traders who get out of all or most of their positions by the end of the day got stuck with tied up capital, interest costs, and overnight risk that needed to be hedged as a result of no fault of their own.
It's not like they "got stuck" with them. They bought them, and hadn't yet sold them.
I have no sympathy.
If you're making stock decisions that rely on 4-minute windows, you should expect to be screwed on a regular basis. Really, trying to time market fluctuations more closely than at least yearly terms--and more realistically, 5 year+ terms--is a crap shoot. Barring highly unusual situations, if a company was a good long-term prospect 4 minutes or 4 days ago then it's still a good prospect now. Really, how many of the positions you held made massive announcements that affected the company's prospects in those 4 minutes?
More to the point, whatever you were holding when trades stopped, you were still holding the next day, and the NYSE going down doesn't affect the value of the companies whose issues you held. Nor was your right to buy or sell shares affected, you just lost 4 minutes of service from the largest market for finding buyers/sellers.
Guys like Warren Buffett--or large fund managers-- don't speculate on minute-to-minute fluctations, and if the market was down for a couple of days it wouldn't really affect their ability to make wise investments. And tiny downtimes like this don't materially affect the fundamentals of the market system.
By nature, if you play with day-trading you're accepting a much higher risk, and when that risk materializes I don't feel that bad for you.
You can have "pretty good" software, you can have "excellent" software, you can even have "never, ever been broken" software, but you can't have "provably perfect" software in all but the most trivial of cases, (and even then things are suspect because of other potential issues: BIOS reliance, hardware manufacturing defects, etc.)
I agree with you on the hardware.
But provable software can exist beyond the trivial. See, e.g., the FOX project at Carnegie Mellon University which has implemented an HTTP server, packet filters, etc in ML, all (mathematically) proven correct.
Proving results about larger software is made much easier by doing strictly functional programming in strong statically typed languages, hence their use of ML.
They've even done research on proof-carrying code, where foreign programs come distributed along with proofs of their behavior that are verified before running them.
Now, there are some caveats:
1. You can prove that certain conditions aren't violated, or that others are always met; whether that spec meets your definition of security is up to you.
2. The FOX verification relies on the correctness of the underlying OS. There's no reason in principle that you can't write a verified OS, but it's not yet been done. (verifying a special-purpose OS may suffice for some purposes).
But the situation is nowhere near as dire as you make it out to be.