Oh, sure, I get that. But what they don't get is that they don't have the leverage to pull it off. They figure that the vaunted New York Times (for instance) has been around forever and Google is some plucky little startup that needs to learn some manners. They don't quite grasp that Google practically owns searching, that that without search, fewer people come to their site.
Basically, they think, mistakenly, that they can dictate terms to Google, and that their news adds more value to Google than Google's free advertising adds to their papers. They couldn't be more wrong, and they're determined to learn the hard way.
To begin, I guess I shouldn't have used the word "frightening", because I've spent enough time around nerds to know that the socially maladjusted are often unable to comprehend anything less than literal. The word used here does not mean "filling the candidate with dread"; it was rather an allusion to the more colloquial (British), "that was a scary test," as in, "that was a tough test." Since a human's physiological response to challenge is often similar to the response associated with fear (cuz makin u a lert iz useful 4 runnin from tigers AND 4 catchin food), the term is not entirely without merit.
The reason I rang up on that wasn't an exercize in pedantry. It's because I've seen far too many profs take pleasure in how much dread and stress they can invoke in their students. I'm from a simpler time, when tests were designed to simply measure what you actually know.
Of course grading is up to the instructor; sometimes the instructor wants to know if you can make a right/wrong decision at speed - he's not giving you an earth-shatteringly complex problem which you're unlikely to solve completely, with the aim of only testing your progress.
I'd still say that short answer would make for a better test in the same situation.
Of course grading is up to the instructor; sometimes the instructor wants to know if you can make a right/wrong decision at speed - he's not giving you an earth-shatteringly complex problem which you're unlikely to solve completely, with the aim of only testing your progress.
True in principle but only for a perfectly constructed multiple choice test, which I've never seen. Also, it yields equal credit for an answer that is very nearly correct with one that is moronically incorrect. If the test contains hundreds of questions, then the differences will average out, but if not, then this becomes meaningful. As a quantum mechanicist, I'll not belabor the point that a few samples may be drastically different from an expectation value.
Even the statistical methods taught in UK high schools are sufficient to rubbish the "it's possible to simply get lucky guessing" argument. In fact, it's a lot easier to analyse than the "do we give him the benefit of the doubt?" or emphasis/stylistic preference/et al. issues that go, say, with essay marking.
As I said, how long are these tests? If they're long enough to get around those sampling problems, then I'd say they're nowhere near complicated enough to sufficiently test proficiency in quantum mechanics. In my quantum class, tests were 4 questions and took a few hours. I'd say a multiple choice test in which question can be answered in about 10 seconds would be a poor measure of actual profiency; rather, it would at best test the ability to memorize.
Whether you personally saw a multiple choice test is neither here nor there, unless you have some supporting background for the decision.
The point is that a good program would never use such a poor device to measure in depth knowledge of a very complicated subject.
But is it really "dumb"? I don't have any doubt that these media outfits have very talented economists, financialists, and lawyers working for them. These are people who can accurately predict what will happen if the media companies were to take this course of action. They know how consumers will respond, and they know how it will affect their company's bottom line.
The question, though, is whether those smart people are actually allowed to make the final decision. This is the newspaper indstry we're talking about. The same industry that is *still* making you start an account and sign in to track what you read. This pisses a lot of people (like me) off, who end up not coming back to the site. Not only that, 90% of that information can be tracked by simply logging IP. So they turn away advertising revenue because of a completely antiquated practice.
So no, it's not surprising that companies like that would fail to figure out that search engines are FREE FUCKING ADVERTISING.
Go check out some of the recent articles on Techdirt, this is one of the author's favorite pet peeves. The upshot is that the newspaper industry has its collective head up its ass and completely fails to understand this whole internet thing. The recent developments in Belgium and France, where newspapers have sued google to avoid being cached, demonstrate this principle in action.
If I'm google, I tell them go ahead - your funeral.
Don't you love knowing that the citizens of other countries, including the People's Republic of China, have more saying in (s)electing their leaders than you do in Yankee Imperialist Bastardistan!?
Good God, keep your trolls less blatant if you want to actually get anybody with them. That was just moronic.
You're an idiot. A well-written multiple choice quiz of the "choose the most appropriate answer" variety for testing particular abilities executed at speed can be made more frightening than a "waffle at length" variety where you can always get something for kinda, sorta, vaguely understanding.
If you equate testing to frightening, then you're worse than an idiot. Grading is up to the instructor, and it is not required to pay by the word. On the other hand, with a "multiple guess" test, it is required that the instructor recognize all possible correct and near correct answers, and it's possible to simply get lucky guessing. So I'll second Gparent, having gone through grad school, I *never* saw a multiple choice test. Ever.
Although if I were going to guess what field would use a multiple choice test with discrete allowed values, it would have to be quantum. Something appropriate about that.
Yes, if you break it down to ASM. But that's kind of why higher level languages exist - to abstractify that shit so it's readable, and to limit the kind of jumps that programmers can make so they make some damned sense. Hopefully.
Library of Alexandria was established by Greek dynasty and has little to do with ancient Egyptian civilization, about which, btw, we indeed knew very little before Napoleon's Egyptian Expedition.
1) The first point is true, but that still does pretty danged far back. Also, much of what was collected at that time did include older knowledge, so the original library would have included some rather old knowledge. 2) The second point depends on the "we." The western world knew much about ancient Egypt at one time, but this knowledge was gradually lost until the 1800s as you point out.
And no, I don't consider zune's wireless sharing to be even the "next little thing." It's like trying to take on Kimberly Clark (maker of Kleenax) by selling your own brand of facial tissue. Sure, you can make money, you might even take some of their market share, but you will never be Kleenax. You have to find a new niche if you want to be the next big thing.
I'd say wireless sharing is much more of a big thing than any other development in mp3 players in the last 5 years, if done right. You know damn well that if Apple does it, people will be going nuts around here.
Caveat: the "play it 3 times" thing isn't the right way.
Unless I'm reading it wrong, the parent poster's point is that until 2000, exit polling did jive with the actual result of the election. After 2000, it did not. Regardless of how flawed exit polls are, the dichotomy indicates a problem unless public behavior radically changed (and I don't think it has).
Unless, maybe, there are other factors at play other than the two you chose? Both those elections were very close, much moreso than usual, so that plays. Close elections are easier to call. Compare that to the elections in 1996,1992,1988,1984,1980,1976, and 1972, all of which could have been predicted by calling about 4 people in each state.
The result is still the same: if the exit poll doesn't jive, figure out how to do a better exit poll.
You can code spaghettic crap in any language. Basic just made it easier which means you were introduced to the reasons *why* you should write plain, procedural modular code at an earlier stage,
Now will their Egyptian counterpart step up and one up them, offering free online access to 3000 years of archived research? Where's the URL for "What the stars look like 180 days before the Nile overflows its banks"?
Unfortunately much of the contents of the library of Alexandria burned centuries ago.
Too many cores on the same bus will cause a lot of contention for memory access. There will always be a place for NUMA architectures, including clusters. That place is for the ultra-high end though, not for scientists who merely want a few processors for a Gaussian computation.
Don't think that would parallelize nicely anyway. Tell ya what, this scientist could use one for quite a few things that *are* nicely parallelizable.
ou can trivally solve that problem by making computers print the finished ballot. A perfectly clear, line by line, printout of who was voted for. You use a standard laser printer, each ballot should fit on a (watermarked, obviously) sheet of paper, because you're just printing the office and candidate who won.(1)
Agreed.
Forget these comparisons to ATMs. It's the same level of complexity as a fucking computerized cash register.
I think you'd be shocked at the number of people who couldn't use a computerized cash register.
I live in Toronto, and the elections held in Canada use paper. Why? Becasuse there's an audit trail if a recount is needed, and it's simple. No duplicated effort. The system isn't broken, and it _just works_
The problem is where analog imperfectly meets digital. Is that a pregnant chad? A hanging chad? A dimpled chad? If the glitches with electronic voting are worked out, it could allow for better error checking of the votes from people who screw up paper ballots.
Personally, I consider the use of paper ballots to be a proper poll test. In other words, if you can't figure that out, you deserve to be disenfranchised.
He challenges other internet companies, including Google, to justify their claim that they could do more good than harm by co-operating with Beijing. Wikipedia has been banned from China since last October
Yeah, I think the second sentence pretty much gives him the answer to the question in the first.
Basically anyone who engages in any serious pirating will cross that $1000 barrier. The issue is how fast, which turns into a matter of who sets the value of the work. If I pirate a song, is the value $10 (for an overpriced CD single in a store) or $0.88 (what I could get it for completely legally from Walmart) or $0.02 (what I could probably get it for quasi-legally on Allofmp3). If they win that argument, you're looking at 100 songs being criminal, possibly as few as 30-40 movies. With software, 2 copies of MS Office would do the trick. So, illegally installing Office on the machines of 2 family members would be a crime, punishable by jail time.
Agree for the most part, except that I severly dislike Konqueror. And yet I use it in fluxbox because Gnome blows goats. Bah.
Incidentally, has Slackware's installer actually changed since 1994? Additionally, am I the only person who likes a Slack install?
Oh, one thing I like about windows: it's a great memory enhancer. If you don't know exactly where to do everything related to system maintenance, you sure as hell aren't finding it. See, Slack is too easy for us old people. Looking for the config file for the **** daemon? Bet it's in/etc/****.conf. That's no fun.
It's OK. Throw it behind a firewall and a router, don't install stupid shit, you'll be OK for the most part. He never said it was wonderful and fantastic. He said it was fine, which is a lukewarm compliment at best.
XP works. It might piss you off. It might be ugly. It might take add-on software to do the things you actually want a computer for. And if you connect it to a bare ethernet port with a real IP address, yeah, it'll get infected. But it's a shitload better than 9x/ME.
To summarize: Great? No. Decent? Sure. Good enough for me? Hell no, not if I can help it.
Slashdot Mirror
Oh, sure, I get that. But what they don't get is that they don't have the leverage to pull it off. They figure that the vaunted New York Times (for instance) has been around forever and Google is some plucky little startup that needs to learn some manners. They don't quite grasp that Google practically owns searching, that that without search, fewer people come to their site.
Basically, they think, mistakenly, that they can dictate terms to Google, and that their news adds more value to Google than Google's free advertising adds to their papers. They couldn't be more wrong, and they're determined to learn the hard way.
Not necessarily. Particularly unlimited use. Hence EULAs.
Come to think of it, I haven't done one for a while...
To begin, I guess I shouldn't have used the word "frightening", because I've spent enough time around nerds to know that the socially maladjusted are often unable to comprehend anything less than literal. The word used here does not mean "filling the candidate with dread"; it was rather an allusion to the more colloquial (British), "that was a scary test," as in, "that was a tough test." Since a human's physiological response to challenge is often similar to the response associated with fear (cuz makin u a lert iz useful 4 runnin from tigers AND 4 catchin food), the term is not entirely without merit.
The reason I rang up on that wasn't an exercize in pedantry. It's because I've seen far too many profs take pleasure in how much dread and stress they can invoke in their students. I'm from a simpler time, when tests were designed to simply measure what you actually know.
Of course grading is up to the instructor; sometimes the instructor wants to know if you can make a right/wrong decision at speed - he's not giving you an earth-shatteringly complex problem which you're unlikely to solve completely, with the aim of only testing your progress.
I'd still say that short answer would make for a better test in the same situation.
Of course grading is up to the instructor; sometimes the instructor wants to know if you can make a right/wrong decision at speed - he's not giving you an earth-shatteringly complex problem which you're unlikely to solve completely, with the aim of only testing your progress.
True in principle but only for a perfectly constructed multiple choice test, which I've never seen. Also, it yields equal credit for an answer that is very nearly correct with one that is moronically incorrect. If the test contains hundreds of questions, then the differences will average out, but if not, then this becomes meaningful. As a quantum mechanicist, I'll not belabor the point that a few samples may be drastically different from an expectation value.
Even the statistical methods taught in UK high schools are sufficient to rubbish the "it's possible to simply get lucky guessing" argument. In fact, it's a lot easier to analyse than the "do we give him the benefit of the doubt?" or emphasis/stylistic preference/et al. issues that go, say, with essay marking.
As I said, how long are these tests? If they're long enough to get around those sampling problems, then I'd say they're nowhere near complicated enough to sufficiently test proficiency in quantum mechanics. In my quantum class, tests were 4 questions and took a few hours. I'd say a multiple choice test in which question can be answered in about 10 seconds would be a poor measure of actual profiency; rather, it would at best test the ability to memorize.
Whether you personally saw a multiple choice test is neither here nor there, unless you have some supporting background for the decision.
The point is that a good program would never use such a poor device to measure in depth knowledge of a very complicated subject.
But is it really "dumb"? I don't have any doubt that these media outfits have very talented economists, financialists, and lawyers working for them. These are people who can accurately predict what will happen if the media companies were to take this course of action. They know how consumers will respond, and they know how it will affect their company's bottom line.
The question, though, is whether those smart people are actually allowed to make the final decision. This is the newspaper indstry we're talking about. The same industry that is *still* making you start an account and sign in to track what you read. This pisses a lot of people (like me) off, who end up not coming back to the site. Not only that, 90% of that information can be tracked by simply logging IP. So they turn away advertising revenue because of a completely antiquated practice.
So no, it's not surprising that companies like that would fail to figure out that search engines are FREE FUCKING ADVERTISING.
Go check out some of the recent articles on Techdirt, this is one of the author's favorite pet peeves. The upshot is that the newspaper industry has its collective head up its ass and completely fails to understand this whole internet thing. The recent developments in Belgium and France, where newspapers have sued google to avoid being cached, demonstrate this principle in action.
If I'm google, I tell them go ahead - your funeral.
Don't you love knowing that the citizens of other countries, including the People's Republic of China, have more saying in (s)electing their leaders than you do in Yankee Imperialist Bastardistan!?
Good God, keep your trolls less blatant if you want to actually get anybody with them. That was just moronic.
You're an idiot. A well-written multiple choice quiz of the "choose the most appropriate answer" variety for testing particular abilities executed at speed can be made more frightening than a "waffle at length" variety where you can always get something for kinda, sorta, vaguely understanding.
If you equate testing to frightening, then you're worse than an idiot. Grading is up to the instructor, and it is not required to pay by the word. On the other hand, with a "multiple guess" test, it is required that the instructor recognize all possible correct and near correct answers, and it's possible to simply get lucky guessing. So I'll second Gparent, having gone through grad school, I *never* saw a multiple choice test. Ever.
Although if I were going to guess what field would use a multiple choice test with discrete allowed values, it would have to be quantum. Something appropriate about that.
I thought a person who acts unsportingly was a cheat. Whereas a cheater's sort of like a leopard
Love that Boston accent:
tuna: A fish. Pronounced like too-ner
tuner: A piece of stereo equipment. Pronounced like too-nah
And they say people in the Appalachians are inbred.
Yes, if you break it down to ASM. But that's kind of why higher level languages exist - to abstractify that shit so it's readable, and to limit the kind of jumps that programmers can make so they make some damned sense. Hopefully.
Library of Alexandria was established by Greek dynasty and has little to do with ancient Egyptian civilization, about which, btw, we indeed knew very little before Napoleon's Egyptian Expedition.
1) The first point is true, but that still does pretty danged far back. Also, much of what was collected at that time did include older knowledge, so the original library would have included some rather old knowledge. 2) The second point depends on the "we." The western world knew much about ancient Egypt at one time, but this knowledge was gradually lost until the 1800s as you point out.
And no, I don't consider zune's wireless sharing to be even the "next little thing." It's like trying to take on Kimberly Clark (maker of Kleenax) by selling your own brand of facial tissue. Sure, you can make money, you might even take some of their market share, but you will never be Kleenax. You have to find a new niche if you want to be the next big thing.
I'd say wireless sharing is much more of a big thing than any other development in mp3 players in the last 5 years, if done right. You know damn well that if Apple does it, people will be going nuts around here.
Caveat: the "play it 3 times" thing isn't the right way.
That should read, close elections are HARDER to call.
Unless I'm reading it wrong, the parent poster's point is that until 2000, exit polling did jive with the actual result of the election. After 2000, it did not. Regardless of how flawed exit polls are, the dichotomy indicates a problem unless public behavior radically changed (and I don't think it has).
Unless, maybe, there are other factors at play other than the two you chose? Both those elections were very close, much moreso than usual, so that plays. Close elections are easier to call. Compare that to the elections in 1996,1992,1988,1984,1980,1976, and 1972, all of which could have been predicted by calling about 4 people in each state.
The result is still the same: if the exit poll doesn't jive, figure out how to do a better exit poll.
Disturbing. Not as disturbing as the hamster thing the previous class did, but still.
You can code spaghettic crap in any language. Basic just made it easier which means you were introduced to the reasons *why* you should write plain, procedural modular code at an earlier stage,
Not without a jump/goto.
Now will their Egyptian counterpart step up and one up them, offering free online access to 3000 years of archived research? Where's the URL for "What the stars look like 180 days before the Nile overflows its banks"?
Unfortunately much of the contents of the library of Alexandria burned centuries ago.
I was coding 1000 line programs in BASIC at age 8,
Of course no one could have debugged that spaghetti crap...
Was that Ch1a? What year?
Too many cores on the same bus will cause a lot of contention for memory access. There will always be a place for NUMA architectures, including clusters. That place is for the ultra-high end though, not for scientists who merely want a few processors for a Gaussian computation.
Don't think that would parallelize nicely anyway. Tell ya what, this scientist could use one for quite a few things that *are* nicely parallelizable.
ou can trivally solve that problem by making computers print the finished ballot. A perfectly clear, line by line, printout of who was voted for. You use a standard laser printer, each ballot should fit on a (watermarked, obviously) sheet of paper, because you're just printing the office and candidate who won.(1)
Agreed.
Forget these comparisons to ATMs. It's the same level of complexity as a fucking computerized cash register.
I think you'd be shocked at the number of people who couldn't use a computerized cash register.
I live in Toronto, and the elections held in Canada use paper. Why? Becasuse there's an audit trail if a recount is needed, and it's simple. No duplicated effort. The system isn't broken, and it _just works_
The problem is where analog imperfectly meets digital. Is that a pregnant chad? A hanging chad? A dimpled chad? If the glitches with electronic voting are worked out, it could allow for better error checking of the votes from people who screw up paper ballots.
Personally, I consider the use of paper ballots to be a proper poll test. In other words, if you can't figure that out, you deserve to be disenfranchised.
He challenges other internet companies, including Google, to justify their claim that they could do more good than harm by co-operating with Beijing. Wikipedia has been banned from China since last October
Yeah, I think the second sentence pretty much gives him the answer to the question in the first.
Basically anyone who engages in any serious pirating will cross that $1000 barrier. The issue is how fast, which turns into a matter of who sets the value of the work. If I pirate a song, is the value $10 (for an overpriced CD single in a store) or $0.88 (what I could get it for completely legally from Walmart) or $0.02 (what I could probably get it for quasi-legally on Allofmp3). If they win that argument, you're looking at 100 songs being criminal, possibly as few as 30-40 movies. With software, 2 copies of MS Office would do the trick. So, illegally installing Office on the machines of 2 family members would be a crime, punishable by jail time.
Agree for the most part, except that I severly dislike Konqueror. And yet I use it in fluxbox because Gnome blows goats. Bah. Incidentally, has Slackware's installer actually changed since 1994? Additionally, am I the only person who likes a Slack install? Oh, one thing I like about windows: it's a great memory enhancer. If you don't know exactly where to do everything related to system maintenance, you sure as hell aren't finding it. See, Slack is too easy for us old people. Looking for the config file for the **** daemon? Bet it's in /etc/****.conf. That's no fun.
It's OK. Throw it behind a firewall and a router, don't install stupid shit, you'll be OK for the most part. He never said it was wonderful and fantastic. He said it was fine, which is a lukewarm compliment at best. XP works. It might piss you off. It might be ugly. It might take add-on software to do the things you actually want a computer for. And if you connect it to a bare ethernet port with a real IP address, yeah, it'll get infected. But it's a shitload better than 9x/ME. To summarize: Great? No. Decent? Sure. Good enough for me? Hell no, not if I can help it.