The original had 4 functions. This one has at least 40. Would you rather the MS guys spend time seeing if they can force their 114k application down into 10k, or perhaps writing an operating system that doesn't suck?
What makes you think I haven't read the rules? I did read them, and downloaded the data, and decided not to participate because I didn't think that the existing algorithms could be usefully improved upon. The results seem to have borne me out.
Perhaps I should have phrased some of this better: what ELSE did they look at and decide against? What did they rent and change their minds about? (That's the second one and I really should have proof-read that better.)
I said in the original thread announcing the competition that I'd rather incorporate more data than trying to get a small improvement (they're only looking for 10%). The contest shows that they haven't missed something obvious, which is useful, but if they're actually after better results rather than validation that they've done an OK job by themselves they'd be better off using data that they have but are currently not included.
They're looking in the wrong places, and trying to squeeze blood from a stone.
User ratings are a deeply flawed way of getting this information. They're one-dimensional and prone to serious randomizations based on the user's mood; a 5 today might have been a 3 tomorrow. Since most of the movies that a user rates will be between 3 and 5 (it's just not that hard to spot a movie you're going to hate, so why would you rent it in the first place?) that makes the data... well, not valueless, but containing a lot less truth than you'd like.
Netflix has a huge amount of additional data that they're not using: * What did the user look at? * What did the user rent? * How did they order their queue? * How long did the user keep the film? * When did the user add additional films that can be considered "related"? * What did the user mark "not interested" (not included in the data set, IIRC).
If they want better recommendations, it's time to stop looking for the quarter under the lamppost and broaden their horizons. You probably can't anonymize all that data well enough to let the world compete for it, but if their internal developers with all that data can't beat outsiders with less, they need to hire some new researchers.
Note to future moderators: "Informative" is for stories containing information. Stories containing obviously untrue things for the purpose of humor are to be modded "Funny" if actually funny, and just ignored otherwise.
You might want to wiki "deadpan humor" if you need a refresher course.
This would be an ugly, ugly case in court. The government will insist at every turn that it would compromise national security to either dismiss the case or deny access to information and people.
So there would be investigations for years, which ultimately would accomplish nothing, all with the goal of possibly punishing a company who will claim that they thought they were doing the patriotic thing.
From a political point of view, Republicans think that they were just defending national security (and therefore deserve to be let off the hook) and some Democrats think that they're going to be portrayed as soft on terrorism.
I'm not saying I'm happy with this reasoning, but that's what's going on in their skulls. Not that you're any happier now that you know it. Sorry.
It's easy to sell things when you're already famous. That's what the record labels do: they make you famous. They put you in record stores and on the radio.
So I don't really care who downloads the albums of famous people. There are plenty of brilliant bands out there who you've never heard of and won't download their albums even when they give them away (and they often do).
Yeah, a bunch of famous people got in the newspaper and made a bunch of money off of it. Big deal.
Spam and junk mail are different problems. The DMA likes junk mail, but hates spam, because it makes it harder for their junk emails to get through.
The DMA really wants to enforce the CAN-SPAM rules that make it easy for smart people to filter out the junk but allows it to sail right through to the dumb people who are more willing to buy their stuff.
The problem is the vast number of spammers who AREN'T part of the DMA. That spam is already illegal under the CAN-SPAM act because of the forged headers, lack of opt-out, etc. They piss you off, but they also piss off the DMA, because the spam filters required to recognize the true spam strip out the legal junk emails as well.
The DMA will enforce the rules on its own members because they're afraid of what spam filters do. If the true spam disappears, you would just filter out the legal spam and there just wouldn't be all that much of it.
Sadly, true spam isn't going away any time soon, so the DMA is pretty much stuck. They should spend some of their money pursuing and prosecuting the illegal spammers, which would make both you and them a lot happier.
"[i]t never happened to anyone who told the truth." That's very odd. A polygraph cuff is just helping to take your blood pressure, and it doesn't hurt. If you set it to the point of pain it wouldn't do any good.
It's not like the thing responds to perceived lies with more pressure, or that the reactions it's measuring are painful. That would completely throw off what little good the polygraph is actually able to do.
So I have no idea why the guy would say that, unless he's not operating the polygraph properly and has no conception of how it's supposed to be used.
Or perhaps he's just trying to throw a scare into the guy. The thing can be set to the point of pain, especially if you put it on wrong with something digging in. I suppose the guy could be threatening to turn it up to 11, which really would hurt a lot, but that would certainly be deliberate torture.
I believe they don't want to push it too hard because easy credit is an important driver in the economy. They give you easy credit, you buy houses and cars and stuff on credit cards, and lots of people get jobs selling you those things.
There's the fact that they make it too easy for people to buy stuff without realizing that they have to pay it back, but it's kind of a separate issue. If they erred on the side of security, the economy would slow drastically. You'd need an economist (which I am not) to run all the numbers, but basically the assertion is that the amount of fraud does less damage to the economy than the good done by easy credit.
What we really need is to make it easy to get credit if you qualify and not if you don't, which means forcing the credit providers to come up with a better mechanism for verifying identity than they're currently using (which is essentially none at all). There are difficulties there with civil liberties, as well as the fact that if you put more faith in a better authentication mechanism you suffer even more when it's broken (and there are no unbreakable authentication mechanisms).
Plus, there's the fact that the credit providers are personally profiting from the current rules. Which means it would be up to government to mandate a better scheme, which (a) they would do badly, like those idiotic RFID passports, and (b) would certainly set records for new forms of civil liberties violations.
I'd say that the real skew on the free market comes from the armies of spambots. They can spew essentially infinite spam, which gives the spammers a huge thumb on the scales in any free competition between us and them.
I'd say the agreement that needs to be made between us is to start shunning ISPs who behave so impolitely. Email is a commons, and subject to the tragedy of the commons. The solution to the tragedy of the commons is politeness.
This commons is so large that there's actually room for considerable freedom, but there are obvious offenders. If you have a machine on your network sending out email 24 hours a day, have a quick peek. If you feel squeamish, make that "peek" an automated spam filter. If it's all spam, SHUT IT DOWN.
Because if you don't, the free market solution is for the rest of us to shut you down.
They're good, but they're not that good for me. I get several spams a day in my inbox (and thousands a day filtered out).
Bizarrely, they should be easy to identify. Most of them are in Russian. Whatever bayesian network they're doing should have figured out by now that I don't read Russian.
The other one is the same template, over and over, all beginning with the same phrase. I have no idea why that one keeps getting through.
I'm sure not complaining; they're clearly filtering out a huge amount of sheer misery.
Just wait. This is Slashdot, where there's at least a vague hope of somebody understanding a bit of science. By the time this hits the regular papers it will be "cuts".
I don't really expect our legal system to have a cogent ordering, where worse crimes do more time. The legal code is a hash composed over decades. I'm sure there are plenty of worse crimes given lesser sentences and vice versa.
Which makes it hard to say exactly what "fit" really means. Jail time serves many different purposes: punishment, vengeance, reformation, deterrence, and simply getting them our of our hair. I can't imagine what it means to optimize for all or any of those things.
Personally, I think that a fairly harsh sentence for spam is appropriate because I want very much for people to stop sending it. I don't know if jail time for spammers will actually achieve that, but I'm so frustrated that I'm willing to give it a try.
It looks like they caught these guys extremely early; many spammers are cranking out that many messages in a few minutes. I can only hope that a few of them will do the math and decide to cut it out.
For my own purposes I'd love to agree. I care far less about MS compatibility than I do about the fact that the interface is awful. So is Word's, and the price is right, so I use it, but I'd really like to see them improve OOo's.
But every time Slashdot mentions OOo, "MS compatibility" is the phrase you hear over and over and over. That office suite, bad as it is (especially in Word), is the de facto standard. When you can't beat the standard when you're giving it away, literally, you know you're stuck.
I just came back from a guitar lesson, and given that I can't seem to hold a beat or finger-pick the same strings on two successive bars, even keeping my fretting hand in my pocket isn't going to make me any good.
The reason is that when miners die in a coal mine collapse, it sucks to be them. When a nuclear plant goes, it sucks to be you.
I'm being flip, but you get the point. A coal mine has zero chance of taking out an entire town. People are concerned that a nuke plant will do what Chernobyl did.
Yes, I know that's never happened in the US. I'm not trying to debate the merits of nuclear power. I'm just explaining why people perceive an actual coal mine disaster as somebody else's problem but a potential nuke plant disaster as their problem.
Like a Google Murder Associates plan? If you need somebody popped, you get to bid on it against other people?
And they'll suggest other people who can also be offed at the same time, perhaps for a discount.
They can make it profitable on both ends, just like with AdWords: not only can you do in the people who need to be done, you can also make a few extra bucks to keep your web site going by taking contract jobs. They provide you with a list of the web sites the inhumee has been browsing lately so you can take a good guess on where to find them, as well as a discount on poisons and ammo purchased with Google Checkout. (Not to mention escape route planning via Google Maps, especially once they add the new Google Cops features to let you know where the police are.)
Last I looked, support in both Word an OO.o for properly laying out illustrations ranged from "intolerably awful" to "criminal neglect". Until this gets better I'm going to cringe in horror at the thought of trying to lay out any document larger than a single page with either program.
The article mentions the ability to put circles and lines on the page but it's so difficult to group them into a single illustration as to make them utterly worthless. I usually end up doing my illustrations in some other program and re-doing the layout every time I change anything.
I'm not talking about trying to mimic Wired Magazine's LSD-driven page layout style. All I want to do is put an image on a page and have it move in a cogent fashion when the text changes, making it look like every single book ever written anywhere.
Interleaf used to do that intelligently, pushing two decades ago now.
Interesting. I take it you've never worked for a small company, or tried to sell something you made.
My background: I run a theater troupe. We make a product that you may want (extremely insightful performances of Shakespeare), but you have many other entertainment options and it may not occur to you to go looking around for live theater (especially amateur live theater) when you can go see a movie, play a video game, etc.
Advertising of some form is the only way we can get an initial audience. Even once you're aware of the existence of my troupe, not everybody will sign up for our mailing list or check the web page often enough to know when performances are. Eventually, with luck, word of mouth spreads, but that takes literally years. Yes, advertising raises my costs, but in fact I can't put on the plays at all without advertising because nobody at all will see them if I don't let people know that they're out there.
I'm simply saying that advertising is not inherently evil. There are bad people who abuse the mechanism, and I keep around a variety of flash blockers and Javascript blockers to get rid of those. But polite advertising may actually show me something I want, and I'd actually prefer to click on the ad to let them know that this is a good form of advertising.
Otherwise you're just ceding the space to those who want to make louder and louder ads with more and more devious ways of bypassing the protection mechanisms.
Some of the spam to a domain I run came to addresses like this:
pydh uuja cmap.ihnc
I've got no idea where these came from. I never used such email addresses anywhere. Oddly, they're used repeatedly; it looks like somebody decided they were valid and then sold them.
Another account had valid addresses "ma" and "ac". I'm quite certain that they couldn't have been harvested, because they were used only in very tightly constrained circumstances. I believe that they were picked randomly. They were replaced by longer email addresses and no spam shows up to those.
And I've seen plenty of emails to "firstname" where that name has never had a valid email address, and it's clear that they're just guessing.
Slashdot Mirror
The original lacked a gui.
And scientific functions.
And the ability to convert hex.
And store/recall.
The original had 4 functions. This one has at least 40. Would you rather the MS guys spend time seeing if they can force their 114k application down into 10k, or perhaps writing an operating system that doesn't suck?
What makes you think I haven't read the rules? I did read them, and downloaded the data, and decided not to participate because I didn't think that the existing algorithms could be usefully improved upon. The results seem to have borne me out.
Perhaps I should have phrased some of this better: what ELSE did they look at and decide against? What did they rent and change their minds about? (That's the second one and I really should have proof-read that better.)
I said in the original thread announcing the competition that I'd rather incorporate more data than trying to get a small improvement (they're only looking for 10%). The contest shows that they haven't missed something obvious, which is useful, but if they're actually after better results rather than validation that they've done an OK job by themselves they'd be better off using data that they have but are currently not included.
They're looking in the wrong places, and trying to squeeze blood from a stone.
User ratings are a deeply flawed way of getting this information. They're one-dimensional and prone to serious randomizations based on the user's mood; a 5 today might have been a 3 tomorrow. Since most of the movies that a user rates will be between 3 and 5 (it's just not that hard to spot a movie you're going to hate, so why would you rent it in the first place?) that makes the data... well, not valueless, but containing a lot less truth than you'd like.
Netflix has a huge amount of additional data that they're not using:
* What did the user look at?
* What did the user rent?
* How did they order their queue?
* How long did the user keep the film?
* When did the user add additional films that can be considered "related"?
* What did the user mark "not interested" (not included in the data set, IIRC).
If they want better recommendations, it's time to stop looking for the quarter under the lamppost and broaden their horizons. You probably can't anonymize all that data well enough to let the world compete for it, but if their internal developers with all that data can't beat outsiders with less, they need to hire some new researchers.
Note to future moderators: "Informative" is for stories containing information. Stories containing obviously untrue things for the purpose of humor are to be modded "Funny" if actually funny, and just ignored otherwise.
You might want to wiki "deadpan humor" if you need a refresher course.
This would be an ugly, ugly case in court. The government will insist at every turn that it would compromise national security to either dismiss the case or deny access to information and people.
So there would be investigations for years, which ultimately would accomplish nothing, all with the goal of possibly punishing a company who will claim that they thought they were doing the patriotic thing.
From a political point of view, Republicans think that they were just defending national security (and therefore deserve to be let off the hook) and some Democrats think that they're going to be portrayed as soft on terrorism.
I'm not saying I'm happy with this reasoning, but that's what's going on in their skulls. Not that you're any happier now that you know it. Sorry.
It's easy to sell things when you're already famous. That's what the record labels do: they make you famous. They put you in record stores and on the radio.
So I don't really care who downloads the albums of famous people. There are plenty of brilliant bands out there who you've never heard of and won't download their albums even when they give them away (and they often do).
Yeah, a bunch of famous people got in the newspaper and made a bunch of money off of it. Big deal.
Spam and junk mail are different problems. The DMA likes junk mail, but hates spam, because it makes it harder for their junk emails to get through.
The DMA really wants to enforce the CAN-SPAM rules that make it easy for smart people to filter out the junk but allows it to sail right through to the dumb people who are more willing to buy their stuff.
The problem is the vast number of spammers who AREN'T part of the DMA. That spam is already illegal under the CAN-SPAM act because of the forged headers, lack of opt-out, etc. They piss you off, but they also piss off the DMA, because the spam filters required to recognize the true spam strip out the legal junk emails as well.
The DMA will enforce the rules on its own members because they're afraid of what spam filters do. If the true spam disappears, you would just filter out the legal spam and there just wouldn't be all that much of it.
Sadly, true spam isn't going away any time soon, so the DMA is pretty much stuck. They should spend some of their money pursuing and prosecuting the illegal spammers, which would make both you and them a lot happier.
It's not like the thing responds to perceived lies with more pressure, or that the reactions it's measuring are painful. That would completely throw off what little good the polygraph is actually able to do.
So I have no idea why the guy would say that, unless he's not operating the polygraph properly and has no conception of how it's supposed to be used.
Or perhaps he's just trying to throw a scare into the guy. The thing can be set to the point of pain, especially if you put it on wrong with something digging in. I suppose the guy could be threatening to turn it up to 11, which really would hurt a lot, but that would certainly be deliberate torture.
I believe they don't want to push it too hard because easy credit is an important driver in the economy. They give you easy credit, you buy houses and cars and stuff on credit cards, and lots of people get jobs selling you those things.
There's the fact that they make it too easy for people to buy stuff without realizing that they have to pay it back, but it's kind of a separate issue. If they erred on the side of security, the economy would slow drastically. You'd need an economist (which I am not) to run all the numbers, but basically the assertion is that the amount of fraud does less damage to the economy than the good done by easy credit.
What we really need is to make it easy to get credit if you qualify and not if you don't, which means forcing the credit providers to come up with a better mechanism for verifying identity than they're currently using (which is essentially none at all). There are difficulties there with civil liberties, as well as the fact that if you put more faith in a better authentication mechanism you suffer even more when it's broken (and there are no unbreakable authentication mechanisms).
Plus, there's the fact that the credit providers are personally profiting from the current rules. Which means it would be up to government to mandate a better scheme, which (a) they would do badly, like those idiotic RFID passports, and (b) would certainly set records for new forms of civil liberties violations.
I'd say that the real skew on the free market comes from the armies of spambots. They can spew essentially infinite spam, which gives the spammers a huge thumb on the scales in any free competition between us and them.
I'd say the agreement that needs to be made between us is to start shunning ISPs who behave so impolitely. Email is a commons, and subject to the tragedy of the commons. The solution to the tragedy of the commons is politeness.
This commons is so large that there's actually room for considerable freedom, but there are obvious offenders. If you have a machine on your network sending out email 24 hours a day, have a quick peek. If you feel squeamish, make that "peek" an automated spam filter. If it's all spam, SHUT IT DOWN.
Because if you don't, the free market solution is for the rest of us to shut you down.
They're good, but they're not that good for me. I get several spams a day in my inbox (and thousands a day filtered out).
Bizarrely, they should be easy to identify. Most of them are in Russian. Whatever bayesian network they're doing should have figured out by now that I don't read Russian.
The other one is the same template, over and over, all beginning with the same phrase. I have no idea why that one keeps getting through.
I'm sure not complaining; they're clearly filtering out a huge amount of sheer misery.
Just wait. This is Slashdot, where there's at least a vague hope of somebody understanding a bit of science. By the time this hits the regular papers it will be "cuts".
I don't really expect our legal system to have a cogent ordering, where worse crimes do more time. The legal code is a hash composed over decades. I'm sure there are plenty of worse crimes given lesser sentences and vice versa.
Which makes it hard to say exactly what "fit" really means. Jail time serves many different purposes: punishment, vengeance, reformation, deterrence, and simply getting them our of our hair. I can't imagine what it means to optimize for all or any of those things.
Personally, I think that a fairly harsh sentence for spam is appropriate because I want very much for people to stop sending it. I don't know if jail time for spammers will actually achieve that, but I'm so frustrated that I'm willing to give it a try.
It looks like they caught these guys extremely early; many spammers are cranking out that many messages in a few minutes. I can only hope that a few of them will do the math and decide to cut it out.
Is that different from the Archimedes Screw, which has been used for well over 2,000 years? It's pretty clever but it's not exactly new.
I don't know, but I know that Dirac is spinning the opposite way.
For my own purposes I'd love to agree. I care far less about MS compatibility than I do about the fact that the interface is awful. So is Word's, and the price is right, so I use it, but I'd really like to see them improve OOo's.
But every time Slashdot mentions OOo, "MS compatibility" is the phrase you hear over and over and over. That office suite, bad as it is (especially in Word), is the de facto standard. When you can't beat the standard when you're giving it away, literally, you know you're stuck.
I just came back from a guitar lesson, and given that I can't seem to hold a beat or finger-pick the same strings on two successive bars, even keeping my fretting hand in my pocket isn't going to make me any good.
*golf clap*
If it does web and email, it has to go out and get connectivity from somewhere. If it uses one of the cellular networks, it might as well be a phone.
If it is just going to use wi-fi, though, it sounds like the new iPod Touch might be pretty close right now.
The reason is that when miners die in a coal mine collapse, it sucks to be them. When a nuclear plant goes, it sucks to be you.
I'm being flip, but you get the point. A coal mine has zero chance of taking out an entire town. People are concerned that a nuke plant will do what Chernobyl did.
Yes, I know that's never happened in the US. I'm not trying to debate the merits of nuclear power. I'm just explaining why people perceive an actual coal mine disaster as somebody else's problem but a potential nuke plant disaster as their problem.
Like a Google Murder Associates plan? If you need somebody popped, you get to bid on it against other people?
And they'll suggest other people who can also be offed at the same time, perhaps for a discount.
They can make it profitable on both ends, just like with AdWords: not only can you do in the people who need to be done, you can also make a few extra bucks to keep your web site going by taking contract jobs. They provide you with a list of the web sites the inhumee has been browsing lately so you can take a good guess on where to find them, as well as a discount on poisons and ammo purchased with Google Checkout. (Not to mention escape route planning via Google Maps, especially once they add the new Google Cops features to let you know where the police are.)
Sign me up.
Last I looked, support in both Word an OO.o for properly laying out illustrations ranged from "intolerably awful" to "criminal neglect". Until this gets better I'm going to cringe in horror at the thought of trying to lay out any document larger than a single page with either program.
The article mentions the ability to put circles and lines on the page but it's so difficult to group them into a single illustration as to make them utterly worthless. I usually end up doing my illustrations in some other program and re-doing the layout every time I change anything.
I'm not talking about trying to mimic Wired Magazine's LSD-driven page layout style. All I want to do is put an image on a page and have it move in a cogent fashion when the text changes, making it look like every single book ever written anywhere.
Interleaf used to do that intelligently, pushing two decades ago now.
Interesting. I take it you've never worked for a small company, or tried to sell something you made.
My background: I run a theater troupe. We make a product that you may want (extremely insightful performances of Shakespeare), but you have many other entertainment options and it may not occur to you to go looking around for live theater (especially amateur live theater) when you can go see a movie, play a video game, etc.
Advertising of some form is the only way we can get an initial audience. Even once you're aware of the existence of my troupe, not everybody will sign up for our mailing list or check the web page often enough to know when performances are. Eventually, with luck, word of mouth spreads, but that takes literally years. Yes, advertising raises my costs, but in fact I can't put on the plays at all without advertising because nobody at all will see them if I don't let people know that they're out there.
I'm simply saying that advertising is not inherently evil. There are bad people who abuse the mechanism, and I keep around a variety of flash blockers and Javascript blockers to get rid of those. But polite advertising may actually show me something I want, and I'd actually prefer to click on the ad to let them know that this is a good form of advertising.
Otherwise you're just ceding the space to those who want to make louder and louder ads with more and more devious ways of bypassing the protection mechanisms.
Is it really surprising that a device that can play videos costs more than a device that just stores them?
Some of the spam to a domain I run came to addresses like this:
pydh
uuja
cmap.ihnc
I've got no idea where these came from. I never used such email addresses anywhere. Oddly, they're used repeatedly; it looks like somebody decided they were valid and then sold them.
Another account had valid addresses "ma" and "ac". I'm quite certain that they couldn't have been harvested, because they were used only in very tightly constrained circumstances. I believe that they were picked randomly. They were replaced by longer email addresses and no spam shows up to those.
And I've seen plenty of emails to "firstname" where that name has never had a valid email address, and it's clear that they're just guessing.