Some of us old-timers remember a time when you could hear your PC on the radio.
My old Apple ][+ jammed my favorite station.
Re:Every toaster on the internet?
on
CNN On IPv6
·
· Score: 3
Every toaster? Nah... Every soda machine, every cash register, every scrolling display sign.
You betchya!
Wouldn't a company love it if they could use ip to tell how full a soda machine was? Wouldn't they love it to change the electronic signs outside their stores?
I used to program cash registers. Do you know how much I would have given to be able to telnet to a misbehaving register on the other side of the country?
There are tremendous business uses for this sort of thing.
After more than 3 years of being a devout Java developer, I've recently switched back to C++. Why? Because Sun's not going to get an screw that up anymore than it already is.
I wasn't really complaining so much as I was curious how all this was supposed to work.
Where it really would get questionable is if Betas reached Windows 2000 like proportions, where hundreds of thousands of people had access to them.
Though I'd think that Corel would be better off letting at least the Beta testers have the source. The type of Beta testers they'll get will likely be the sorts who would give them a file name and line number of the problem. (However, one could argue that this sort of beta tester doesn't make for a particularly good study for the intended target user of this thing.)
The biggest problem with most neural nets is that they are inherently parallel algorithms implemented on a serial architecture. As such, they are horrendously slow. I think you'll find that they start to do useful things once implemented on truly parallel hardware.
I did once attempted to implement a very simple short-term memory system using a neural net when in college. In doing so, I managed to use up twice the cpu quota for the entire class in one evening on a Vax. At that point, it was decided that the assignment did not need completing.
According to their FAQ, they won't release source code until they make their final release. Doesn't that violate the GPL? I thought you always had to release source if you released binaries...
Re:Having everyone always moderate is dangerous
on
Moderation Ideas
·
· Score: 1
That would work, but it seems somewhat inferior to the way it works now. All I think that it would do would be to cause ratings to fluctuate wildly depending on who was on. That doesn't seem very conducive to pointing out better things to read. If the above happens too often, people are just going to give up on the whole concept.
I think that no matter what you do, moderation is going to take a bit of work. After all, you are supposedly thinking about the article before moderating, right?
Having everyone always moderate is dangerous
on
Moderation Ideas
·
· Score: 3
Solution: Allow all readers to be moderators all of the time. There will never be a deficiency in the number of moderators with this method. The number of people scoring articles will be directly proportional to the number of people reading that article. When 100% of the participants can cast votes, there is no outlying possibility if lack of moderation.
The trouble with this is that most people will soon cease bothering to moderate. When it is a rare occurance, people will take it as an honor, and work to do a good job. When it is a constant duty, people will slack off and not moderate. The only people who will moderate are those with very strong opinions on the subject, and I'd suggest that they are the least likely to be objective.
If you take KDE vs. Gnome as an example, with random moderation, you have to assume that a large percentage of the moderators will be in that largest class of people who doesn't greatly care one way or another and moderation will be fair. If everyone has the power to moderate, most people (the moderate moderators) will be sick of moderating and probably won't bother. The only ones who will are those who think it is really important to moderate posts up or down. Those are likely to be the more fanatical posters, and you won't get objective moderation.
Moderation will just turn into a new sort of internet poll, and we all know how accurate those are!
Ideally, you'd always want moderators who had no opinion on the subject. Since that isn't likely possible, choosing at random is likely the next best bet.
I do like your other ideas, though. The averaging of scores would help one problem, which is that in popular topics you seen pages and pages of fours and fives while in a lesser used topic, you might see only a couple of twos and threes. Obviously this has little to do with post quality!
I just got done reading The Wild Shore by C S Friedman that explores a number of themes related to this. Without going into too much detail, it describes an off-shoot of the human race that has a much wider range of psychological differences the we do. Different people have different types of disorders which, rather than being cured, are used. One of the main characters, a computer security expert, has a "disorder" that is very similar to autism and that allows him to better do his job.
(It is also a pretty good post-cyberpunk story, for what it's worth.)
Why a different distributions?
on
Linux Lite?
·
· Score: 1
I don't see why different distributions are needed to accomplish this. It seems that the same thing could be accomplished with two options presented near the beginning, "Express" and "Advanced". "Express" would create the system this guy is talking about. "Advanced" would give you all the options. It should be doable with one package.
If hackers were truly communist, they would choose to write programs "for the greater good". Instead, most tend to write programs they are interested in. In this, open source programming is no less a selfish activity then programming for a "capitalist" boss. The only difference is that in one the coin of the realm is cool software while in the other it is cash.
People work with systems like Linux because for most of them, this is the easiest they can work on on the sorts of projects they want to work on. If they love what they do, this, in and of itself, is the motivating factor. Money doesn't enter into it.
The giving away of the software afterwords is a sort of global deal that allows hackers to work on such projects. No single hacker could build an OS. A bunch of them together can. If your goal is to build an OS, and you don't have a job at Microsoft, pretty much the only way is to share code with your buddies.
That's what drives open source, not altruism. And in that, it is as fully "greed" oriented as capitalism. Which is, of course, why it works while communist systems, relying on altruism, mostly fail.
In software there is no cost to the owner to give it away. People like the above author don't understand this and thus confuse it for communism.
That's what I mean. We give up security for convenience.
Personally, I'd take the inconvenience of having to be issued a new card when moving if it meant that the card would be utterly and completely useless if stolen.
But then, some people complain when the cashier wants to see an ID with a credit card. And then they wonder how the guy who stole their card managed to charge it to the limit without getting caught.
One of the reasons that banks love these things (and why consumer acceptence might be slow) is that it passes off much of the liability to the consumer. Because the cash is actually "in" the card, the card becomes identical to cash. If the card is stolen, the card-holder is out of luck. As long as the account data is stored on the physical card, this pretty much has to be the case.
In contrast, if a debit card or credit card is stolen, the card-holder is typically liable only for $50. (Or $0 if they can show that the retailers using the card did not make an appropriate security effort, which they almost never do. Generally, if the signatures don't match your ID, you aren't out anything.)
It should be obvious why the banks would love to see everyone walking around with smart cards instead of credit cards. Their own liability goes down. The same goes for retailers. They barely check IDs now, can you imagine how little they'll do it if they have no liability in accepting stolen cards?
A similar thing happens with those "gift certificate" cards you can get at blockbuster video. These aren't typically smart cards, however, it is interesting that in most cases, name data is not stored either on the card or at the host where the data is kept. The reason for this is deliberate. Companies don't want to deal with lost cards and the like. By refusing to store name data, the card becomes just like cash, and all card security is the consumer's problem.
However, it would be nice if they offered the customers an option that would make it so that their card could only be used if it was swiped (I honestly don't know how they would do this, I'm just hypothesizing).
I work in the retail EFT industry. Currently, nearly all protocols aimed at retail transactions capture this information ("swiped" vs. "keyed"). Internet based protocols usually don't, as nearly nobody has a card reader by their machine, but one imagines that could be changed fairly quickly.
Of course, the problem that you are going to have is that someone who knows these protocols could fake it pretty easily.
I doubt they actually do this, but one very good way to prevent the above would be to embed the shipping address in the card and then refuse to ship elsewhere.
Unfortunately, that would probably fail because of "inconvenience", the bane of all security.
Though on the flip side of the coin, without using anything but your peripheral vision, try to count the number (or even color) of major items on the desk in front of you.
Though if they are in a pattern, you don't, which is interesting in and of itself. We don't have to count the dots every time a die comes up six.
Much of this may have to do with the fact that human vision is not like a computer's vision. We don't get a nice rectangle of even pixels. We see best in a small area where are vision is focused. Move a few degrees out and vision (peripheral vision) rapidly degrades. This alone means that it is next to impossible to examine two things well at the same time.
I'd be very curious as to if they'd get similar results with sound experiments. I suspect not. I recall experiments in which people were fed different sound sources in each ear. Even when paying attention to one voice stream for some task, subjects still responded to their own names in the other voice stream. This implies that some level of cognitive processing is still going on for the stream supposedly being ignored.
This isn't all that surprising. It has long been known that trying to do two "similar" tasks is very difficult. It is very hard to sing while reading, for example, so in that sense, the brain does have units that are serial. However, there is quite a bit of parallism in that you can fairly easily do disimilar tasks. Examining blocks while singing, for example. In that respect, the brain is somewhat like a modern CPU. A small set of discrete, basically serial devices that can operate in parallel. (A gross oversimplification, but a good analogy, I think.)
At a low level (at the neuron level) we already know that the brain has to be massively parallel. The best proof of this is that it takes about 100 milliseconds for a neuron to respond to a signal. Obviously you couldn't do much with a serial algorithm at that speed.
Yes, competition is good. It is only unfair competition that is a bad thing.
The advantage of open source is that no one can win, at least not in the sort of way Microsoft can. Even if Red Hat gained 90% of the market, they still wouldn't have the sort of monopoly Microsoft does. The barriers to entry in competing directly with Microsoft are huge. You have to put together the capital to build a new OS. The barriers to entry in competing directly with a theoretical Red Hat monopoly are very small. All you have to do is copy some CDs, hire some good people and you're there.
This, of course, is a hugely good thing, as it means that regardless of market shares, Linux companies are going to be under competitive pressure to improve.
Slashdot Mirror
Some of us old-timers remember a time when you could hear your PC on the radio.
My old Apple ][+ jammed my favorite station.
Every toaster? Nah... Every soda machine, every cash register, every scrolling display sign.
You betchya!
Wouldn't a company love it if they could use ip to tell how full a soda machine was? Wouldn't they love it to change the electronic signs outside their stores?
I used to program cash registers. Do you know how much I would have given to be able to telnet to a misbehaving register on the other side of the country?
There are tremendous business uses for this sort of thing.
After more than 3 years of being a devout Java developer, I've recently switched back to C++. Why? Because Sun's not going to get an screw that up anymore than it already is.
yeah, that's Microsoft's job!
- A disgruntled VC++ developer
I wasn't really complaining so much as I was curious how all this was supposed to work.
Where it really would get questionable is if Betas reached Windows 2000 like proportions, where hundreds of thousands of people had access to them.
Though I'd think that Corel would be better off letting at least the Beta testers have the source. The type of Beta testers they'll get will likely be the sorts who would give them a file name and line number of the problem. (However, one could argue that this sort of beta tester doesn't make for a particularly good study for the intended target user of this thing.)
The biggest problem with most neural nets is that they are inherently parallel algorithms implemented on a serial architecture. As such, they are horrendously slow. I think you'll find that they start to do useful things once implemented on truly parallel hardware.
I did once attempted to implement a very simple short-term memory system using a neural net when in college. In doing so, I managed to use up twice the cpu quota for the entire class in one evening on a Vax. At that point, it was decided that the assignment did not need completing.
Those seem to be their generic Beta requirements. Note the references to Macs below it. I bet no one remembered to change the page.
According to their FAQ, they won't release source code until they make their final release. Doesn't that violate the GPL? I thought you always had to release source if you released binaries...
That would work, but it seems somewhat inferior to the way it works now. All I think that it would do would be to cause ratings to fluctuate wildly depending on who was on. That doesn't seem very conducive to pointing out better things to read. If the above happens too often, people are just going to give up on the whole concept.
I think that no matter what you do, moderation is going to take a bit of work. After all, you are supposedly thinking about the article before moderating, right?
Solution: Allow all readers to be moderators all of the time. There will never be a deficiency in the number of moderators with this method. The number of people scoring articles will be directly proportional to the number of people reading that article. When 100% of the participants can cast votes, there is no outlying possibility if lack of moderation.
The trouble with this is that most people will soon cease bothering to moderate. When it is a rare occurance, people will take it as an honor, and work to do a good job. When it is a constant duty, people will slack off and not moderate. The only people who will moderate are those with very strong opinions on the subject, and I'd suggest that they are the least likely to be objective.
If you take KDE vs. Gnome as an example, with random moderation, you have to assume that a large percentage of the moderators will be in that largest class of people who doesn't greatly care one way or another and moderation will be fair. If everyone has the power to moderate, most people (the moderate moderators) will be sick of moderating and probably won't bother. The only ones who will are those who think it is really important to moderate posts up or down. Those are likely to be the more fanatical posters, and you won't get objective moderation.
Moderation will just turn into a new sort of internet poll, and we all know how accurate those are!
Ideally, you'd always want moderators who had no opinion on the subject. Since that isn't likely possible, choosing at random is likely the next best bet.
I do like your other ideas, though. The averaging of scores would help one problem, which is that in popular topics you seen pages and pages of fours and fives while in a lesser used topic, you might see only a couple of twos and threes. Obviously this has little to do with post quality!
(It is also a pretty good post-cyberpunk story, for what it's worth.)
I don't see why different distributions are needed to accomplish this. It seems that the same thing could be accomplished with two options presented near the beginning, "Express" and "Advanced". "Express" would create the system this guy is talking about. "Advanced" would give you all the options. It should be doable with one package.
Just slap a thumbnail of Michelangelo's "David" and an excerpt from "Lolita" on every page.
Keep in mind that bacteria aren't the simplest forms of life known.
A virus is far simpler, to the point where scientists still argue about whether or not they are alive.
If isn't really "From each according to his ability..." because developers develop what they want to rather than making the best use of their ability.
If hackers were truly communist, they would choose to write programs "for the greater good". Instead, most tend to write programs they are interested in. In this, open source programming is no less a selfish activity then programming for a "capitalist" boss. The only difference is that in one the coin of the realm is cool software while in the other it is cash.
People work with systems like Linux because for most of them, this is the easiest they can work on on the sorts of projects they want to work on. If they love what they do, this, in and of itself, is the motivating factor. Money doesn't enter into it.
The giving away of the software afterwords is a sort of global deal that allows hackers to work on such projects. No single hacker could build an OS. A bunch of them together can. If your goal is to build an OS, and you don't have a job at Microsoft, pretty much the only way is to share code with your buddies.
That's what drives open source, not altruism. And in that, it is as fully "greed" oriented as capitalism. Which is, of course, why it works while communist systems, relying on altruism, mostly fail.
In software there is no cost to the owner to give it away. People like the above author don't understand this and thus confuse it for communism.
That's what I mean. We give up security for convenience.
Personally, I'd take the inconvenience of having to be issued a new card when moving if it meant that the card would be utterly and completely useless if stolen.
But then, some people complain when the cashier wants to see an ID with a credit card. And then they wonder how the guy who stole their card managed to charge it to the limit without getting caught.
How come? Because "ignorant executive" is redundant.
One of the reasons that banks love these things (and why consumer acceptence might be slow) is that it passes off much of the liability to the consumer. Because the cash is actually "in" the card, the card becomes identical to cash. If the card is stolen, the card-holder is out of luck. As long as the account data is stored on the physical card, this pretty much has to be the case.
In contrast, if a debit card or credit card is stolen, the card-holder is typically liable only for $50. (Or $0 if they can show that the retailers using the card did not make an appropriate security effort, which they almost never do. Generally, if the signatures don't match your ID, you aren't out anything.)
It should be obvious why the banks would love to see everyone walking around with smart cards instead of credit cards. Their own liability goes down. The same goes for retailers. They barely check IDs now, can you imagine how little they'll do it if they have no liability in accepting stolen cards?
A similar thing happens with those "gift certificate" cards you can get at blockbuster video. These aren't typically smart cards, however, it is interesting that in most cases, name data is not stored either on the card or at the host where the data is kept. The reason for this is deliberate. Companies don't want to deal with lost cards and the like. By refusing to store name data, the card becomes just like cash, and all card security is the consumer's problem.
However, it would be nice if they offered the customers an option that would make it so that their card could only be used if it was swiped (I honestly don't know how they would do this, I'm just hypothesizing).
I work in the retail EFT industry. Currently, nearly all protocols aimed at retail transactions capture this information ("swiped" vs. "keyed"). Internet based protocols usually don't, as nearly nobody has a card reader by their machine, but one imagines that could be changed fairly quickly.
Of course, the problem that you are going to have is that someone who knows these protocols could fake it pretty easily.
I doubt they actually do this, but one very good way to prevent the above would be to embed the shipping address in the card and then refuse to ship elsewhere.
Unfortunately, that would probably fail because of "inconvenience", the bane of all security.
But obviously something is doing enough processing to tell the difference between your name and some other random word.
Though on the flip side of the coin, without using anything but your peripheral vision, try to count the number (or even color) of major items on the desk in front of you.
Though if they are in a pattern, you don't, which is interesting in and of itself. We don't have to count the dots every time a die comes up six.
Much of this may have to do with the fact that human vision is not like a computer's vision. We don't get a nice rectangle of even pixels. We see best in a small area where are vision is focused. Move a few degrees out and vision (peripheral vision) rapidly degrades. This alone means that it is next to impossible to examine two things well at the same time.
I'd be very curious as to if they'd get similar results with sound experiments. I suspect not. I recall experiments in which people were fed different sound sources in each ear. Even when paying attention to one voice stream for some task, subjects still responded to their own names in the other voice stream. This implies that some level of cognitive processing is still going on for the stream supposedly being ignored.
This isn't all that surprising. It has long been known that trying to do two "similar" tasks is very difficult. It is very hard to sing while reading, for example, so in that sense, the brain does have units that are serial. However, there is quite a bit of parallism in that you can fairly easily do disimilar tasks. Examining blocks while singing, for example. In that respect, the brain is somewhat like a modern CPU. A small set of discrete, basically serial devices that can operate in parallel. (A gross oversimplification, but a good analogy, I think.)
At a low level (at the neuron level) we already know that the brain has to be massively parallel. The best proof of this is that it takes about 100 milliseconds for a neuron to respond to a signal. Obviously you couldn't do much with a serial algorithm at that speed.
Yes, competition is good. It is only unfair competition that is a bad thing.
The advantage of open source is that no one can win, at least not in the sort of way Microsoft can. Even if Red Hat gained 90% of the market, they still wouldn't have the sort of monopoly Microsoft does. The barriers to entry in competing directly with Microsoft are huge. You have to put together the capital to build a new OS. The barriers to entry in competing directly with a theoretical Red Hat monopoly are very small. All you have to do is copy some CDs, hire some good people and you're there.
This, of course, is a hugely good thing, as it means that regardless of market shares, Linux companies are going to be under competitive pressure to improve.