Eh? What part of "Even in translations that don't feature lots of 'thou shalt not' and 'thus spake'" did you not understand? While using the KJV certainly enhances the problem, even the more colloquial translations use a very different sort of English than that used for nearly anything else.
The Bible is an excellent and important resource for getting started on MT (machine translation) and NLP (natural language processing) in general, especially for languages with smaller speaker bases.
The problem is, the language style used is very specific to the Bible. Even in translations that don't feature lots of "thou shalt not" and "thus spake", you get some really strange constructions that make it unsuitable for most tasks, unless you really have nothing else to work with.
IP is definitely considered in these things. While you could probably obtain an illicit copy of most of the text corpora out there just to play around with it, if you ever intend to publish your research, you need to "buy" the corpus, which gives you the right to use it to build translation systems, parsers, question-answerers, or whatever else.
Typically, academic licences are a lot cheaper than commercial ones, although the base price can vary all over the place. The Canadian Hansards (parliamentary proceedings, in French and English---a major corpus used in statistical machine translation work, including Och's) will run you $5k; the ECI Multilingual Corpus 1 is about $35. Usually, corpora are made available through either the Linguistic Data Consortium in Philadelphia or the Evaluations and Language resources Distribution Agency in Paris, although some of the free corpora are distributed elsewhere, typically from the website of the research lab that developed it.
Two major costs go into the creation of corpora: content and markup. The former is often responsible for the majority of the cost, as LDC or ELDA negotiate with the copyright holder for a redistribution licence, although the markup costs can be significant for more a elabourately-annotated corpus, such as a treebank (which contains parse structure and more for all the sentences in it). However, assuming you can get enough free content, or negotiate for a free licence to the content, there's no theoretical reason there couldn't be an open corpus repository....
You're absolutely right---if we're talking about a finished system. For research purposes, we need to find out just how far a "dumb" system will take us; and then we can add in the domain knowledge later. If we can get a "dumb" system to outperform one with lots of domain knowledge, then A) it is _clear_ that it is the algorithmic framework of the dumb system that is better, and not just that it had better domain experts, B) if that's the level of performance you want, why waste time with the domain experts?, and C) odds are good that you can improve performance even more by integrating domain knowledge into the dumb system, if you can only figure out how.
That last part is important. More often than not, it seems, enhancing a totally stupid statistical algorithm with domain knowledge actually *hurts* at first, until you figure out just the right way to provide the knowledge to the system without breaking the statistical algorithms. This is an active and very interesting area of research....
Yup, that's right---the Brown et al paper (and a lot of the work since then) was on translating between French and English; these were chosen because the Canadian parliamentary proceedings (known as the Canadian Hansards---"hansard" being a general term for parliamentary proceedings within the British Commonwealth) are in both languages. Oddly enough, the corpus isn't as gargantuan as you might imagine; in order to be useful to these systems, it needs to be *at least* sentence-aligned (i.e. "this English sentence corresponds to this French sentence") and preferably word-aligned as well, but of course the originals aren't, and this needs to be done by hand.
What they are is expensive. You can order them from the Linguistic Data Consortium, but it'll run you a cool $5K---and that's if you're doing academic, non-commercial research. The nice thing about Natural Language Processing, though, is that that's the *only* real expense beyond a medium-powerful computer; no special hardware required.:)
Different forms of medical devices have different requirements for precision---that's the percent of things they report that are correct---and recall---that's the percent of things they are supposed to find, that they do find. Before this could replace an MRI as a primary diagnostic tool, it would need to have 100% recall. Precision is less important, because if you do get a false positive, you can go in for further testing (which is not to diminish the stress and fear induced by false positives).
In the article the experts comment that it can't replace MRIs yet with the reported accuracy rate, which is true. What it *can* do is become a routine part of a physical. Once the wand is bought, it sounds like using it is essentially free (well, they have to occasionally replace the batteries!), so there's no reason not to just use it all over the place. And what *that* does is make it more likely that we'll detect tumours early, before we would have any reason to suspect a problem otherwise.
On the problem of false positives, btw---a lot of that can be mitigated by a good doctor. When a false positive is possible, the doctor needs to explain that, exaggerating its possibility, to reduce the stress in the meantime. My mom went through that once; it turned out that she was just unusually athletic and the more muscular tissue set off the sensor! But those three days sucked. Had the doctor said, "look, this is probably nothing, we just need to run another test", it's not nearly as bad.
Two funny (sad) arithmetic stories
on
Making Change
·
· Score: 4, Funny
A few years back, my dad was paying for something, and paid an uneven amount in order to get even change. The clerk looked at the money, sort of shrugged, and punched it in and started counting out the change. The catch is---my dad misheard the amount. So when the clerk started counting out a bunch of pennies and nickels, my dad was like, "wait, what?" Had the clerk had *any idea* why my dad had given an uneven amount, she would have realised that he'd misheard the price. But she just punched it in and started counting it out....
A few years after that, my sister (in 5th grade at the time) had a test with a miscalculated grade, and when my mom went in for a parent-teacher conference, she brought it up. In particular, she said she'd added up the number correct and divided by the total number of questions, and got a different percentage... the teacher looked down her nose at my mom and said, "that's *not* how it's calculated." How was it calculated? Well, you have these cardboard discs that you turn according to the total number of questions, and then you read the grade out of the little window corresponding to the number right.... This woman had only the vaguest notion that this grade was a percentage correct, and *no idea at all* that---as a percentage---it could also be calculated by dividing the numbers out. None.
Correct me if I'm wrong, but that's going to put two semicolons after the if---not itself an error, but it's going to block the else from working properly, since the second semicolon is a(n empty) statement after the end of the if clause!
In general, using macros in this fashion is dangerous and difficult to maintain, because it's not known what exactly is in the content of the macro. If a shortcut is at least part of a language, then a maintainer has some reasonable chance of understanding a piece of code as written without going hunting through header files. (Though of course one of the major complaints against perl is that there are so many shortcuts that are part of the language that nobody knows them all, pretty much reducing us to the previous problem.)
I see ++foo all the time, without any special meaning. In fact, I've conditioned myself never to write foo++ unless I need to. For one thing, there's the performance issue if foo is an object---no need to keep a copy with ++foo. But more importantly, the semantics of ++foo are much more in keeping with other expressions in the language. If I write
dfoo = ++foo * 2;
then it works the same as if I'd written
dfoo = (foo += 1) * 2;
or
dfoo = (foo = foo + 1) * 2;
---not that either of those is any better, but in all of them, there is a sense of "do everything for this operator, reduce it to a value, then plug into the surrounding context". By contrast, for
dfoo = foo++ * 2;
you can't rewrite the "foo++" as anything else while preserving semantics, without spreading it over multiple statements.
So I take issue with your comment that programmers do, or even should, use the postfix form by default. Not that it really matters; well-written code ends up using this operator so very rarely that it doesn't end up mattering very often.:)
Whenever the topic of internet-based voting comes up, there's always a lot of fuss over security; what people seem to miss is that security isn't the problem: secrecy is.
Folks, you don't want an individual vote to be checkable. If there is any way for me to prove conclusively---to the gov't or anyone else---how I voted, then there is likewise a way for someone to pay me or force me to vote a certain way! Mandatory secrecy is crucial. Whether curtains are involved or not, poll volunteers are supposed to be careful not to let people show their votes (accidentally or intentionally) to anyone else.
This doesn't exclude using computer terminals at the polling station to actually improve the voting user interface, which I wholly support. But making people physically go to the polling place is the only way I know of to guarantee secrecy, so I'm very much against internet voting.
Yeah, I've been watching on Univision, mostly because I don't get ESPN; but after I'd already seen the US-Poland game, I caught the last ten minutes of it on ABC two days later (*two days* delay, jeez), and they had something like two basketball analogies and one (American) football analogy just in the time I saw it. And then they went ON and ON and ON about how completely bizarre it was that we moved on to the elimination rounds in the "strangest possible way"---you know, by doing better than the other teams. But the commentators just didn't know the rules or something, and were really weirded out by the fact that we could lose the last game of the group and still take second within the group.
Anyway, yeah, I've been watching Univision. The only thing I miss is the explanations for the fouls ("una falta! mumble mumble mumble...") and cards. Better than just watching on mute, though, because I can get the excitement level from the commentators at least. And I swear I understand more Spanish than I did three weeks ago.:)
Re:What a crappy review (spoilers rot13d)
on
Minority Report
·
· Score: 1
There's yet another explanation:
Vg'f ng yrnfg cbffvoyr gung jung jr fnj vf whfg gur fgnoyr ybbcvat irefvba, gur "svkrq-cbvag" vs lbh jvyy, naq gung gur "bevtvany" jnl gung riragf hasbyqrq vf gung nsgre frggvat hc gur thl jvgu cvpgherf bs Frna, Ohetrff qebccrq n uvag gb Naqregba, be neenatrq sbe uvz gb svaq n yrnq, be jungrire. Naqregba sbyybjf hc, svaqf guvf thl, xvyyf uvz. Nu, ohg gur cer-pbtf frr gung, erq-onyy gurve cerqvpgvba, naq va gur arj havirefr gung'f npghnyyl gur gvcbss. Ohg gur obbgfgenc qvqa'g arrq gb pbzr sebz gur cer-pbtf, cre fr.
This can't be repeated often enough. This site cannot possibly hope to be taken seriously when they put Barbarella on this list. I mean, Barbarella. Hello? Plausibility? Futurism? Entertainment, I suppose, in that MST3K sort of way. Come on.
Coins will probably be phased out (if inflation hits again and no soda costs less than a dollar, which you already see in high profile areas with vending machines), but dollar bills should be available for more than your life-time.
Not likely. Vending machine people hate dollar bills, because they're harder to authenticate; people who use vending machines don't like them much either, given the likelihood of any one bill failing to be accepted (come on, tell me you haven't stood in front of a vending machine with just one dollar bill on your person, feverishly trying to straighten it out so the damn machine will take it).
The US govt has issued a new dollar coin, and while its use isn't yet widespread, it is on the rise. In particular, more and more vending machines accept them, and various banks are now pushing their use. I suspect that while the treasury won't stop printing dollar bills, they'll trail them off to slowly force people into using coins. If nothing else, coins are much cheaper in the long run, since they last on the order of ten times as long as bills.
Go ahead, look in your wallet---I bet you don't see any bills in there older than 1995 or so. Unless something strange happens, like a bill getting stuffed in a drawer and forgotten, the life of a printed note is just a few years. By contrast, if you look at the coins in your pocket, chances are quite good that you have at least a few coins dating from the 60s. This goes for any foreign country that doesn't have an inflation economy, as well! Coins are much sturdier, therefore cheaper in the long run, therefore preferable to the government....
Hint: Just because you've had it pounded into your brain that socialism is somehow evil, doesn't make it true.
Naah, I happily consider myself to be socialist (my party affiliation is Green: definitely a variety of socialist). I was just pointing out that, while the idea isn't what you'd usually call libertarian, it's compatible with their worldview of "let the market work it out" by using taxes to incorporate a number of market externalities directly into the market---perfecting the market, as it were. I don't doubt that there exist libertarians that would oppose this plan, but there are a number of libertarians I know that are in favour of it as well.
Since you are down on "liberal/socialists", who is going to pay for the mass transit systems you want? Clearly the market will not provide.
Well, the market might provide (I believe would, but we can't know for sure until we try it) if we actually made the market internalise all of its costs. That is, driving a car does damage to roads and the environment, as well as creating noise pollution and creating traffic that slows down the other cars; but none of these costs are factored in to the driving of the car (except possibly the first: some states funnel parts of their gas taxes into road repair, though that is rarely the only source of such funding). As a result, driving a car seems much cheaper.
It is usually considered a socialist (or "European", which seems to mean the same thing these days to Americans, *sigh*) idea to increase gas taxes and car taxes, but it actually is fully compatible with a libertarian view as well, particularly as it means that the market actually can work out the best balance between driving cars and taking mass transit, not to mention providing a source of funding for things like road building and repair, noise and exhaust barrier construction, and environmental cleanup, so that it's the people who use a thing that pay for it.
So to answer your question: raise gas, new car, and annual car taxes and the market will provide.
Watch out for that (Verizon) SingleRate. It says "no roaming" but in the fine print you find
that this means no roaming charges WITHIN THEIR NETWORK. Furthermore, to get
no-roaming, you need to buy a tri-mode phone.
I have Verizon's SingleRate plan, and was told that it was no roaming charges and no long-distance charges anywhere; and in fact this seems to be the case. I've made calls from a bunch of places and never got charged extra (or indeed at all) for them. Also, the phones that come with this plan (an AudioVox or a StarTac) are trimode.
That said, I'm not 100% sure whether I'd recommend Verizon. Overall I've had a fairly good experience, but out of the five or so bills I've gotten to date, two of them have been off by a few cents (80 cents in their favour one month, 13 in my favour the next), which doesn't really bode well for their accounting skills.:( On the other hand, with the error in their favour, when I called up they did fix it, so who knows. Nowadays there are a number of single-rate plans, though, so shop around.
They got this stuff from OpenStep. Some of the old OpenStep software companies (like Stone Design) offered bounties (like free software) for people that translated their apps to other languages.
What are you talking about? This has been an ability of Mac applications for a very long time, courtesy careful use of the resource fork. The strings were kept in a resource, and to translate to a different language, you just pop open ResEdit and edit the relevant strings.
Are resource forks (or something similar) going to still be a part of MacOS X, or will this wonderful feature be lost in the name of "compatibility with foreign OSes"?
Go ahead and run the numbers. See how much the government takes from you *then* ash who is being greedy.
...but first, compare to taxation in every other first world country, and witness how little it actually is. And while you're at it, consider how many services you receive for that 55%....
how might consumption taxes be designed so that they weren't regressive?
They are made non-regressive by combining them with other taxes that are non-regressive. For instance, a progressive income tax. I've heard it said that in many areas of the country the combination of income, sales, and property tax conspire to make the overall level of taxation roughly the same percentage for everyone. IANAEconomist, so I certainly couldn't say whether that's true---but it is plausible, and if it is not already the case, it could be made to be the case with some judicious legislation.
Quoth Katz: ``Under the law, anyone who makes, sells or uses a device -- software, hardware, or a computer -- that makes copyright circumvention possible is engaging in a criminal act. This is the reason downloading free music and sharing Napster sites had been curtailed on college campuses in recent weeks.''
Bzzt, wrong. For many (most?) of the colleges who banned Napster, the reason was network bandwidth, pure and simple. When a single program ends up taking half to three fourths of the available network bandwidth, performance of other network applications suffers noticeably. Ergo, the application gets banned. (And here at Brown, the ban resulted a sudden and substantial improvement of network performance that I noticed and commented on before even being aware of the reason.)
Of course, if this net hog were some vital academic application, then the computer services departments might work out some other solution; because mp3 trading is illegal, they aren't going to bust their butts to reinstate service. But it should be very clear that its illegality was not in most places the reasons for it being banned.
Something I just don't understand here. Why should the networks care about the rebroadcast? They make money off of advertising; as a result, if their viewer base increases (and they have some way to monitor it), then their ad revenues should increase. If they could get iCraveTV to send them viewership numbers (to forward to their advertisers), then it seems to me this would be a good deal.
This reminds me of something my hometown radio station did... they started webcasting a few months ago, but introduced a 45 second delay so that people wouldn't be able to call in on the contests. Why not? Isn't the purpose of the contest to get more people to listen in? And since they know how many people are listening to the webcast (and can pass it on to their advertisers, same story as above), isn't that all good for them?
Something I just don't understand here. Why should the networks care about the rebroadcast? They make money off of advertising; as a result, if their viewer base increases (and they have some way to monitor it), then their ad revenues should increase. If they could get iCraveTV to send them viewership numbers (to forward to their advertisers), then it seems to me this would be a *good* deal.
This reminds me of something my hometown radio station did... they started webcasting a few months ago, but introduced a 45 second delay so that people wouldn't be able to call in on the contests. Why not? Isn't the purpose of the contest to get more people to listen in? And since they know how many people are listening to the webcast (and can pass it on to their advertisers, same story as above), isn't that all good for them?
Sure, go ahead! But before you do, you should do a few other things:
First, tidy up your code. Fix the indentations, elabourate the variable names, make it generally more readable.
Then, comment it well.
That done, write out your maintainer-level documentation, which includes all of your high-level algorithms.
Verify that these algorithms are correct.
Check the complexity of these algorithms. The real place to shave off time is if you can lower your running time by an O(n) or two.
Once you've fixed your algorithms, update your documentation.
And now, after you've done every other improvement you can think of... save a copy of your code.
Run a profiler on your code.
If you really still feel like it's too slow, then go ahead and optimise it in the places the profiler picks out as taking a lot of time.
The upshot is, optimising is never a bad thing. It's just an extremely low priority thing. I can't tell you how many of my students I've seen performing low-level optimisations on O(n^3) loops that could have been reduced to O(n lg n) or less. Don't waste your time optimising in the wrong places, and really, make sure you've done everything else first....
Slashdot Mirror
Eh? What part of "Even in translations that don't feature lots of 'thou shalt not' and 'thus spake'" did you not understand? While using the KJV certainly enhances the problem, even the more colloquial translations use a very different sort of English than that used for nearly anything else.
The Bible is an excellent and important resource for getting started on MT (machine translation) and NLP (natural language processing) in general, especially for languages with smaller speaker bases.
The problem is, the language style used is very specific to the Bible. Even in translations that don't feature lots of "thou shalt not" and "thus spake", you get some really strange constructions that make it unsuitable for most tasks, unless you really have nothing else to work with.
IP is definitely considered in these things. While you could probably obtain an illicit copy of most of the text corpora out there just to play around with it, if you ever intend to publish your research, you need to "buy" the corpus, which gives you the right to use it to build translation systems, parsers, question-answerers, or whatever else.
Typically, academic licences are a lot cheaper than commercial ones, although the base price can vary all over the place. The Canadian Hansards (parliamentary proceedings, in French and English---a major corpus used in statistical machine translation work, including Och's) will run you $5k; the ECI Multilingual Corpus 1 is about $35. Usually, corpora are made available through either the Linguistic Data Consortium in Philadelphia or the Evaluations and Language resources Distribution Agency in Paris, although some of the free corpora are distributed elsewhere, typically from the website of the research lab that developed it.
Two major costs go into the creation of corpora: content and markup. The former is often responsible for the majority of the cost, as LDC or ELDA negotiate with the copyright holder for a redistribution licence, although the markup costs can be significant for more a elabourately-annotated corpus, such as a treebank (which contains parse structure and more for all the sentences in it). However, assuming you can get enough free content, or negotiate for a free licence to the content, there's no theoretical reason there couldn't be an open corpus repository....
You're absolutely right---if we're talking about a finished system. For research purposes, we need to find out just how far a "dumb" system will take us; and then we can add in the domain knowledge later. If we can get a "dumb" system to outperform one with lots of domain knowledge, then A) it is _clear_ that it is the algorithmic framework of the dumb system that is better, and not just that it had better domain experts, B) if that's the level of performance you want, why waste time with the domain experts?, and C) odds are good that you can improve performance even more by integrating domain knowledge into the dumb system, if you can only figure out how.
That last part is important. More often than not, it seems, enhancing a totally stupid statistical algorithm with domain knowledge actually *hurts* at first, until you figure out just the right way to provide the knowledge to the system without breaking the statistical algorithms. This is an active and very interesting area of research....
Yup, that's right---the Brown et al paper (and a lot of the work since then) was on translating between French and English; these were chosen because the Canadian parliamentary proceedings (known as the Canadian Hansards---"hansard" being a general term for parliamentary proceedings within the British Commonwealth) are in both languages. Oddly enough, the corpus isn't as gargantuan as you might imagine; in order to be useful to these systems, it needs to be *at least* sentence-aligned (i.e. "this English sentence corresponds to this French sentence") and preferably word-aligned as well, but of course the originals aren't, and this needs to be done by hand.
:)
What they are is expensive. You can order them from the Linguistic Data Consortium, but it'll run you a cool $5K---and that's if you're doing academic, non-commercial research. The nice thing about Natural Language Processing, though, is that that's the *only* real expense beyond a medium-powerful computer; no special hardware required.
Different forms of medical devices have different requirements for precision---that's the percent of things they report that are correct---and recall---that's the percent of things they are supposed to find, that they do find. Before this could replace an MRI as a primary diagnostic tool, it would need to have 100% recall. Precision is less important, because if you do get a false positive, you can go in for further testing (which is not to diminish the stress and fear induced by false positives).
In the article the experts comment that it can't replace MRIs yet with the reported accuracy rate, which is true. What it *can* do is become a routine part of a physical. Once the wand is bought, it sounds like using it is essentially free (well, they have to occasionally replace the batteries!), so there's no reason not to just use it all over the place. And what *that* does is make it more likely that we'll detect tumours early, before we would have any reason to suspect a problem otherwise.
On the problem of false positives, btw---a lot of that can be mitigated by a good doctor. When a false positive is possible, the doctor needs to explain that, exaggerating its possibility, to reduce the stress in the meantime. My mom went through that once; it turned out that she was just unusually athletic and the more muscular tissue set off the sensor! But those three days sucked. Had the doctor said, "look, this is probably nothing, we just need to run another test", it's not nearly as bad.
A few years back, my dad was paying for something, and paid an uneven amount in order to get even change. The clerk looked at the money, sort of shrugged, and punched it in and started counting out the change. The catch is---my dad misheard the amount. So when the clerk started counting out a bunch of pennies and nickels, my dad was like, "wait, what?" Had the clerk had *any idea* why my dad had given an uneven amount, she would have realised that he'd misheard the price. But she just punched it in and started counting it out....
A few years after that, my sister (in 5th grade at the time) had a test with a miscalculated grade, and when my mom went in for a parent-teacher conference, she brought it up. In particular, she said she'd added up the number correct and divided by the total number of questions, and got a different percentage... the teacher looked down her nose at my mom and said, "that's *not* how it's calculated." How was it calculated? Well, you have these cardboard discs that you turn according to the total number of questions, and then you read the grade out of the little window corresponding to the number right.... This woman had only the vaguest notion that this grade was a percentage correct, and *no idea at all* that---as a percentage---it could also be calculated by dividing the numbers out. None.
Your answer highlights the very reason that C macros are a bad idea. You say:
#define NOTHING ;
...
if ((flag & VAL1) || (flag & VAL2)
NOTHING;
Correct me if I'm wrong, but that's going to put two semicolons after the if---not itself an error, but it's going to block the else from working properly, since the second semicolon is a(n empty) statement after the end of the if clause!
In general, using macros in this fashion is dangerous and difficult to maintain, because it's not known what exactly is in the content of the macro. If a shortcut is at least part of a language, then a maintainer has some reasonable chance of understanding a piece of code as written without going hunting through header files. (Though of course one of the major complaints against perl is that there are so many shortcuts that are part of the language that nobody knows them all, pretty much reducing us to the previous problem.)
I see ++foo all the time, without any special meaning. In fact, I've conditioned myself never to write foo++ unless I need to. For one thing, there's the performance issue if foo is an object---no need to keep a copy with ++foo. But more importantly, the semantics of ++foo are much more in keeping with other expressions in the language. If I write
dfoo = ++foo * 2;
then it works the same as if I'd written
dfoo = (foo += 1) * 2;
or
dfoo = (foo = foo + 1) * 2;
---not that either of those is any better, but in all of them, there is a sense of "do everything for this operator, reduce it to a value, then plug into the surrounding context". By contrast, for
dfoo = foo++ * 2;
you can't rewrite the "foo++" as anything else while preserving semantics, without spreading it over multiple statements.
So I take issue with your comment that programmers do, or even should, use the postfix form by default. Not that it really matters; well-written code ends up using this operator so very rarely that it doesn't end up mattering very often. :)
Whenever the topic of internet-based voting comes up, there's always a lot of fuss over security; what people seem to miss is that security isn't the problem: secrecy is.
Folks, you don't want an individual vote to be checkable. If there is any way for me to prove conclusively---to the gov't or anyone else---how I voted, then there is likewise a way for someone to pay me or force me to vote a certain way! Mandatory secrecy is crucial. Whether curtains are involved or not, poll volunteers are supposed to be careful not to let people show their votes (accidentally or intentionally) to anyone else.
This doesn't exclude using computer terminals at the polling station to actually improve the voting user interface, which I wholly support. But making people physically go to the polling place is the only way I know of to guarantee secrecy, so I'm very much against internet voting.
Yeah, I've been watching on Univision, mostly because I don't get ESPN; but after I'd already seen the US-Poland game, I caught the last ten minutes of it on ABC two days later (*two days* delay, jeez), and they had something like two basketball analogies and one (American) football analogy just in the time I saw it. And then they went ON and ON and ON about how completely bizarre it was that we moved on to the elimination rounds in the "strangest possible way"---you know, by doing better than the other teams. But the commentators just didn't know the rules or something, and were really weirded out by the fact that we could lose the last game of the group and still take second within the group.
:)
Anyway, yeah, I've been watching Univision. The only thing I miss is the explanations for the fouls ("una falta! mumble mumble mumble...") and cards. Better than just watching on mute, though, because I can get the excitement level from the commentators at least. And I swear I understand more Spanish than I did three weeks ago.
There's yet another explanation:
Vg'f ng yrnfg cbffvoyr gung jung jr fnj vf whfg gur fgnoyr ybbcvat irefvba, gur "svkrq-cbvag" vs lbh jvyy, naq gung gur "bevtvany" jnl gung riragf hasbyqrq vf gung nsgre frggvat hc gur thl jvgu cvpgherf bs Frna, Ohetrff qebccrq n uvag gb Naqregba, be neenatrq sbe uvz gb svaq n yrnq, be jungrire. Naqregba sbyybjf hc, svaqf guvf thl, xvyyf uvz. Nu, ohg gur cer-pbtf frr gung, erq-onyy gurve cerqvpgvba, naq va gur arj havirefr gung'f npghnyyl gur gvcbss. Ohg gur obbgfgenc qvqa'g arrq gb pbzr sebz gur cer-pbtf, cre fr.
This can't be repeated often enough. This site cannot possibly hope to be taken seriously when they put Barbarella on this list. I mean, Barbarella. Hello? Plausibility? Futurism? Entertainment, I suppose, in that MST3K sort of way. Come on.
Jesus fuck, Barbarella! What were they thinking?
The US govt has issued a new dollar coin, and while its use isn't yet widespread, it is on the rise. In particular, more and more vending machines accept them, and various banks are now pushing their use. I suspect that while the treasury won't stop printing dollar bills, they'll trail them off to slowly force people into using coins. If nothing else, coins are much cheaper in the long run, since they last on the order of ten times as long as bills.
Go ahead, look in your wallet---I bet you don't see any bills in there older than 1995 or so. Unless something strange happens, like a bill getting stuffed in a drawer and forgotten, the life of a printed note is just a few years. By contrast, if you look at the coins in your pocket, chances are quite good that you have at least a few coins dating from the 60s. This goes for any foreign country that doesn't have an inflation economy, as well! Coins are much sturdier, therefore cheaper in the long run, therefore preferable to the government....
Naah, I happily consider myself to be socialist (my party affiliation is Green: definitely a variety of socialist). I was just pointing out that, while the idea isn't what you'd usually call libertarian, it's compatible with their worldview of "let the market work it out" by using taxes to incorporate a number of market externalities directly into the market---perfecting the market, as it were. I don't doubt that there exist libertarians that would oppose this plan, but there are a number of libertarians I know that are in favour of it as well.
Well, the market might provide (I believe would, but we can't know for sure until we try it) if we actually made the market internalise all of its costs. That is, driving a car does damage to roads and the environment, as well as creating noise pollution and creating traffic that slows down the other cars; but none of these costs are factored in to the driving of the car (except possibly the first: some states funnel parts of their gas taxes into road repair, though that is rarely the only source of such funding). As a result, driving a car seems much cheaper.
It is usually considered a socialist (or "European", which seems to mean the same thing these days to Americans, *sigh*) idea to increase gas taxes and car taxes, but it actually is fully compatible with a libertarian view as well, particularly as it means that the market actually can work out the best balance between driving cars and taking mass transit, not to mention providing a source of funding for things like road building and repair, noise and exhaust barrier construction, and environmental cleanup, so that it's the people who use a thing that pay for it.
So to answer your question: raise gas, new car, and annual car taxes and the market will provide.
I have Verizon's SingleRate plan, and was told that it was no roaming charges and no long-distance charges anywhere; and in fact this seems to be the case. I've made calls from a bunch of places and never got charged extra (or indeed at all) for them. Also, the phones that come with this plan (an AudioVox or a StarTac) are trimode.
That said, I'm not 100% sure whether I'd recommend Verizon. Overall I've had a fairly good experience, but out of the five or so bills I've gotten to date, two of them have been off by a few cents (80 cents in their favour one month, 13 in my favour the next), which doesn't really bode well for their accounting skills. :( On the other hand, with the error in their favour, when I called up they did fix it, so who knows. Nowadays there are a number of single-rate plans, though, so shop around.
They got this stuff from OpenStep. Some of the old OpenStep software companies (like Stone Design) offered bounties (like free software) for people that translated their apps to other languages.
What are you talking about? This has been an ability of Mac applications for a very long time, courtesy careful use of the resource fork. The strings were kept in a resource, and to translate to a different language, you just pop open ResEdit and edit the relevant strings.
Are resource forks (or something similar) going to still be a part of MacOS X, or will this wonderful feature be lost in the name of "compatibility with foreign OSes"?
Here's a map of Nunavut, Canada, with Devon Island highlighted, in case you're wondering where exactly it is.
...but first, compare to taxation in every other first world country, and witness how little it actually is. And while you're at it, consider how many services you receive for that 55%....
how might consumption taxes be designed so that they weren't regressive?
They are made non-regressive by combining them with other taxes that are non-regressive. For instance, a progressive income tax. I've heard it said that in many areas of the country the combination of income, sales, and property tax conspire to make the overall level of taxation roughly the same percentage for everyone. IANAEconomist, so I certainly couldn't say whether that's true---but it is plausible, and if it is not already the case, it could be made to be the case with some judicious legislation.
Quoth Katz: ``Under the law, anyone who makes, sells or uses a device -- software, hardware, or a computer -- that makes copyright circumvention possible is engaging in a criminal act. This is the reason downloading free music and sharing Napster sites had been curtailed on college campuses in recent weeks.''
Bzzt, wrong. For many (most?) of the colleges who banned Napster, the reason was network bandwidth, pure and simple. When a single program ends up taking half to three fourths of the available network bandwidth, performance of other network applications suffers noticeably. Ergo, the application gets banned. (And here at Brown, the ban resulted a sudden and substantial improvement of network performance that I noticed and commented on before even being aware of the reason.)
Of course, if this net hog were some vital academic application, then the computer services departments might work out some other solution; because mp3 trading is illegal, they aren't going to bust their butts to reinstate service. But it should be very clear that its illegality was not in most places the reasons for it being banned.
Something I just don't understand here. Why should the networks care about the rebroadcast? They make money off of advertising; as a result, if their viewer base increases (and they have some way to monitor it), then their ad revenues should increase. If they could get iCraveTV to send them viewership numbers (to forward to their advertisers), then it seems to me this would be a good deal.
This reminds me of something my hometown radio station did... they started webcasting a few months ago, but introduced a 45 second delay so that people wouldn't be able to call in on the contests. Why not? Isn't the purpose of the contest to get more people to listen in? And since they know how many people are listening to the webcast (and can pass it on to their advertisers, same story as above), isn't that all good for them?
Something I just don't understand here. Why should the networks care about the rebroadcast? They make money off of advertising; as a result, if their viewer base increases (and they have some way to monitor it), then their ad revenues should increase. If they could get iCraveTV to send them viewership numbers (to forward to their advertisers), then it seems to me this would be a *good* deal.
This reminds me of something my hometown radio station did... they started webcasting a few months ago, but introduced a 45 second delay so that people wouldn't be able to call in on the contests. Why not? Isn't the purpose of the contest to get more people to listen in? And since they know how many people are listening to the webcast (and can pass it on to their advertisers, same story as above), isn't that all good for them?
Sure, go ahead! But before you do, you should do a few other things:
- First, tidy up your code. Fix the indentations, elabourate the variable names, make it generally more readable.
- Then, comment it well.
- That done, write out your maintainer-level documentation, which includes all of your high-level algorithms.
- Verify that these algorithms are correct.
- Check the complexity of these algorithms. The real place to shave off time is if you can lower your running time by an O(n) or two.
- Once you've fixed your algorithms, update your documentation.
- And now, after you've done every other improvement you can think of... save a copy of your code.
- Run a profiler on your code.
- If you really still feel like it's too slow, then go ahead and optimise it in the places the profiler picks out as taking a lot of time.
The upshot is, optimising is never a bad thing. It's just an extremely low priority thing. I can't tell you how many of my students I've seen performing low-level optimisations on O(n^3) loops that could have been reduced to O(n lg n) or less. Don't waste your time optimising in the wrong places, and really, make sure you've done everything else first....