She should at least lose her teaching license. Regardless of how the porn got onto the computer, her handling of the situation after that was grossly incompetent for a teacher. She could have simply covered the monitor. An elementary school classroom would have plenty of items available to allow that (construction paper, tape). She could have sent the kids to the playground or cafeteria or assembly room. She could have effectively sought help. (She did seek help, but all she said was there was a pop-up problem. She did not say that they were inappropriate pop-ups). If there was no place to send the kids and she did not want to leave them along while seeking help, she could have brough them with her.
Many stories on the net portray this as if there were some pop-ups, and a couple kids got a glance at them, briefly. In fact, the pop-ups were appearing for several hours, and the kids were exposed to them for several hours. That in all that time she could not come up with a single effective way to deal with the situation shows that she is totally unable to deal with being in charge of children. I'd hate to see how she would do in a medical emergency or a natural disaster or a fire.
100% certainty that 10 people sample-set is too little for a Yes-No experiement
Really? We are testing the hypothesis that people can tell 128k and 256k apart. If the hypothesis is false, then it will be 50/50 whether they get one right or not. The chances of getting 8 or more right out of 10 when an individual trial has probability 1/2 is C(10,8)(1/2)^8(1/2)^2 + C(10,9)(1/2)^9(1/2) + C(10,10)(1/2)^10. That's 56/1024, or 6.3%. That's pretty good grounds for rejecting the hypothesis.
For the other test, using the Shure headphones, where 6 or more got it right, that happens 36.7% of the time when p=1/2. That's not good enough to reject the hypothesis.
Whether 10 is a big enough sample for an experiment like this can only be determined after you run the experiment, and see the results. If the results are close to 5, then either probably either it wasn't big enough or people can't tell them apart. But when it is far from 5 (and 8 is far from 5), then it WAS enough for a trial.
If my name and itunes account info start showing up on music all over P2P sites, the evil RIAA may come knocking on my door.
Using P2P sites is not the usual mechanism one uses to share music with friends, so this is not much of a problem. I share music with friends, for example, by loading the songs onto their iPods. Even the iTunes music that has DRM allows that.
The thing is, the constitution is clear on this. The states don't have a right to charge taxes on stuff shipped across state lines. Why are we even having this discussion?
We're having this discussion because the Constitution doesn't say what you seem to think it says.
I understand that OLPC got the ball rolling, but that doesn't make it the best solution. If its about charity, then grandstanding shouldn't get in the way
Indeed. If you go read the literature at the OLPC website, you'll notice something. You won't be able to tell if the OLPC is the best solution, because there isn't an actual problem they are addressing! Basically, their idea is that if they make a cool, cheap, machine and give it to a bunch of third-world kids, something good might happen. In a large part, it is a big social experiment.
Then compare to the Classmate project. There, you find a project where they have talked to people involved currently in third-world education, such as teachers, and found out what they actually want and need, and tried to give them that.
It's very simple for Microsoft to deal with this. Let's say Linux goes GPLv3, and SuSE is on 10.7 by then, which is the last GPLv2 version. SuSE 10.8 is GPLv3.
You get one of the coupons. You wait a couple years, and by then the current SuSE is 11.2.
You turn in your coupon.
And guess what? Microsoft or Novell or whoever handles fulfilling the coupons sends you a bright new shiny copy of SuSE 10.7.
We once ran a small chat and gaming system where I worked. It had a general chat channel, and when people paired up to play a game, such as chess or checkers, it would make a private channel for them.
Because this was a family service, we had to try to police conduct in the general channel, and because we didn't have the staff to monitor it live 24/7, it fell to me to try to automate some of this. That actually worked fairly well. We had a very large dictionary of naughty words and phrases. When you said something, my filters basically looked for any of those things, and '*'ed them out. The filter ignored whitespace, and it also considered certain characters to be equivalent, so if you wrote 5h17, that would match 'shit', since it knew a 5 could take the place of an s, and so on. However, before filtering, it did a spell check on your text, and marked all the words that were spelled right and were not on the bad word list as safe. For example, if you said "wash it", it would not see the "sh it" as something bad.
This worked surprisingly well. It caught it when people tried tricks like inserting spaces to break up the bad words, but usually did not get false positives, because of the spell check protector stuff. Well, unless you were a lousy speller, but if a lousy speller got kicked off incorrectly for profanity, it still improved things.:-)
One other little trick it did. When it filtered out something in your message, it only did that on the message sent to other people. The copy that echoed back to your system was uncensored.
When you got caught, it would send you a message warning you to watch your language. If you ignored the warning, an admin bot would ban you for a period of time. Repeared bans would be for longer times.
One thing that disappointed me: no one ever tried to use Klingon profanity to get around the filters. I had that covered in the filters, and was hoping to see the reaction when the users discovered that.
Yes. We learned that an AI solution is not necessarily the best way to deal with an AI problem. Deep Blue was an engineering solution, not an AI solution.
You can see a similar thing in Google's automatic language translation. It's purely statistical--they look at a lot of bilingual texts (such as minutes of the UN), and develop a statistical model to come up with translations of new documents. There's no attempt to build any AI into this. It's just a statistics problem and a data structures problem to them, and the result is the best automatic translation out there.
My guess is that many current challenging AI problems will eventually be solved with methods like this, that don't have anything in them that resemble what we think of as "intelligence".
Microsoft clearly can make good set-top box software. Their UltimateTV product was very nice. I had it with DirecTV. When I first got it, it was a little behind Tivo in some features, and ahead in others. First update, a few months after I got it, put it ahead of Tivo on most features, and next year's update made it even better than that.
Then Microsoft shut down the UTV group, transfering the people off to, I think, the XBox group, apparently planning to someday integrate games and set-top boxes.
The only problem UTV had was that the interface was a little slow. But they overcame much of that with good interface design. E.g., the buttons on a page might have been slower than the equivalent on my current Comcast box, but the UTV interface only required a couple button presses to accomplish the task, whereas the Comcast box requires about 5 times as many. Two slow buttons are a lot faster than 10 moderately responsive buttons!
It is completely irrelevant to me what US laws Pirate Bay broke. I support them because they are in Sweden.[...]Why can the MPAA coerce the Swedish police to conduct raids in accordance with US copyright law?
I take it that it never occured to you that Sweden has copyright laws? As does nearly every other country on Earth?
What I do is default to my local Wal-Mart Supercenter (which also happens to be the 2nd nearest store, so is most convenient). On average, their prices easily beat Safeway, Albertson's, and the small regional chain store that is a little closer to me than the Wal-Mart. Furthermore, they are indeed "everyday low prices". This lets me easily memorize Wal-Mart's prices for most of my regular items.
If I happen to be shopping at one of the other places, it is then easy to see if their current sale price beats Wal-Mart's everyday price. If so, and I need the item, or can reasonably store it, I will buy at the other store. If I notice an advertised sale on something at one of the other stores, I might go there for it.
OK, just what is the problem here? Someone slips something into a bill at the last minute, and someone else gets it taken out, and now people think getting it taken out is shady.
OK, let's break it down. Broadcom already has a fully-functional driver for their chipset.
The people who did the Linux driver reverse engineered that in order to figure out how to use the chipset. The Linux people then wrote a driver.
There are some areas where they (the Linux driver writers) think that their code is more clever than the Broadcom code.
They (the Linux driver writers) think Broadcom would want to replace the working code it already has, and that its programmers are familiar with, and that (from a user point of view) does everything the Linux code does.
That seems unlikely to me. Replacing working code with outside code, which probably won't fit into your existing code without extensive reworking, for the sole benefit of making your code more elegant, is simply not something companies are interested in doing.
According to that post and the things it links to, they are afraid Broadcom might want their implementation of something Broadcom already has code for. Basically, they think there code is more clever.
The likelihood that Broadcom would have any interest in the kind of extensive work it would take to rewrite their own code to incorporate that part of the Linux driver, for no benefit whatsoever to the end user, is very low.
Apparently the Linux kernel developer did not wish Broadcom to take advantage of his work in proprietary products
Just how likely is it that there is anything in the Linux driver that would be useful to Broadcom? Broadcom already has fully functional proprietary drivers for their chips.
Suppose you add artificial neurons to a brain, and remove natural neurons as they die. Eventually, you end up with your mind running entirely on artificial neurons. Is your mind now effectively immortal?
And is it still you? Or a copy of you running on the artificial brain? If it is a copy, when does it cease to be you?
A use tax is not a tax or duty laid on exports from another state. It's a tax on having or using items--their origin is irrelevant.
For example, if a state wanted to have a $50/year tax on all automobiles in the state, there would be no problem under Article I, Section 9. They would not have to limit the tax to automobiles that were purchased in the state.
Slashdot Mirror
Many stories on the net portray this as if there were some pop-ups, and a couple kids got a glance at them, briefly. In fact, the pop-ups were appearing for several hours, and the kids were exposed to them for several hours. That in all that time she could not come up with a single effective way to deal with the situation shows that she is totally unable to deal with being in charge of children. I'd hate to see how she would do in a medical emergency or a natural disaster or a fire.
Well, glibc is not under GPLv2, so it seems unlikely to go GPLv3. It is under LGPL.
Well, Science does accept LaTex, according to the submission guidelines. And they then run it through a DOS program that converts LaTeX to Word.
Really? We are testing the hypothesis that people can tell 128k and 256k apart. If the hypothesis is false, then it will be 50/50 whether they get one right or not. The chances of getting 8 or more right out of 10 when an individual trial has probability 1/2 is C(10,8)(1/2)^8(1/2)^2 + C(10,9)(1/2)^9(1/2) + C(10,10)(1/2)^10. That's 56/1024, or 6.3%. That's pretty good grounds for rejecting the hypothesis.
For the other test, using the Shure headphones, where 6 or more got it right, that happens 36.7% of the time when p=1/2. That's not good enough to reject the hypothesis.
Whether 10 is a big enough sample for an experiment like this can only be determined after you run the experiment, and see the results. If the results are close to 5, then either probably either it wasn't big enough or people can't tell them apart. But when it is far from 5 (and 8 is far from 5), then it WAS enough for a trial.
Selling used CDs is legal. It is covered under 17 USC 109. It is no different than selling used books.
Using P2P sites is not the usual mechanism one uses to share music with friends, so this is not much of a problem. I share music with friends, for example, by loading the songs onto their iPods. Even the iTunes music that has DRM allows that.
We're having this discussion because the Constitution doesn't say what you seem to think it says.
Indeed. If you go read the literature at the OLPC website, you'll notice something. You won't be able to tell if the OLPC is the best solution, because there isn't an actual problem they are addressing! Basically, their idea is that if they make a cool, cheap, machine and give it to a bunch of third-world kids, something good might happen. In a large part, it is a big social experiment.
Then compare to the Classmate project. There, you find a project where they have talked to people involved currently in third-world education, such as teachers, and found out what they actually want and need, and tried to give them that.
You get one of the coupons. You wait a couple years, and by then the current SuSE is 11.2.
You turn in your coupon.
And guess what? Microsoft or Novell or whoever handles fulfilling the coupons sends you a bright new shiny copy of SuSE 10.7.
Except the notion of "effectively distributed" is one that does not appear in the law.
Because this was a family service, we had to try to police conduct in the general channel, and because we didn't have the staff to monitor it live 24/7, it fell to me to try to automate some of this. That actually worked fairly well. We had a very large dictionary of naughty words and phrases. When you said something, my filters basically looked for any of those things, and '*'ed them out. The filter ignored whitespace, and it also considered certain characters to be equivalent, so if you wrote 5h17, that would match 'shit', since it knew a 5 could take the place of an s, and so on. However, before filtering, it did a spell check on your text, and marked all the words that were spelled right and were not on the bad word list as safe. For example, if you said "wash it", it would not see the "sh it" as something bad.
This worked surprisingly well. It caught it when people tried tricks like inserting spaces to break up the bad words, but usually did not get false positives, because of the spell check protector stuff. Well, unless you were a lousy speller, but if a lousy speller got kicked off incorrectly for profanity, it still improved things. :-)
One other little trick it did. When it filtered out something in your message, it only did that on the message sent to other people. The copy that echoed back to your system was uncensored.
When you got caught, it would send you a message warning you to watch your language. If you ignored the warning, an admin bot would ban you for a period of time. Repeared bans would be for longer times.
One thing that disappointed me: no one ever tried to use Klingon profanity to get around the filters. I had that covered in the filters, and was hoping to see the reaction when the users discovered that.
The Legend of Zelda: A Pain in my Ass (Part 1)
The Legend of Zelda: A Pain in my Ass (Part 2)
You can see a similar thing in Google's automatic language translation. It's purely statistical--they look at a lot of bilingual texts (such as minutes of the UN), and develop a statistical model to come up with translations of new documents. There's no attempt to build any AI into this. It's just a statistics problem and a data structures problem to them, and the result is the best automatic translation out there.
My guess is that many current challenging AI problems will eventually be solved with methods like this, that don't have anything in them that resemble what we think of as "intelligence".
Then Microsoft shut down the UTV group, transfering the people off to, I think, the XBox group, apparently planning to someday integrate games and set-top boxes.
The only problem UTV had was that the interface was a little slow. But they overcame much of that with good interface design. E.g., the buttons on a page might have been slower than the equivalent on my current Comcast box, but the UTV interface only required a couple button presses to accomplish the task, whereas the Comcast box requires about 5 times as many. Two slow buttons are a lot faster than 10 moderately responsive buttons!
Sweden is dealing with it.
You seem to think it odd that when someone in Sweden allegedly breaks Swedish law, they are prosecuted in Sweden by a Swedish prosecutor.
I take it that it never occured to you that Sweden has copyright laws? As does nearly every other country on Earth?
If I happen to be shopping at one of the other places, it is then easy to see if their current sale price beats Wal-Mart's everyday price. If so, and I need the item, or can reasonably store it, I will buy at the other store. If I notice an advertised sale on something at one of the other stores, I might go there for it.
Well, slipping it in was just as shady.
What ever gave you that idea?
The people who did the Linux driver reverse engineered that in order to figure out how to use the chipset. The Linux people then wrote a driver.
There are some areas where they (the Linux driver writers) think that their code is more clever than the Broadcom code.
They (the Linux driver writers) think Broadcom would want to replace the working code it already has, and that its programmers are familiar with, and that (from a user point of view) does everything the Linux code does.
That seems unlikely to me. Replacing working code with outside code, which probably won't fit into your existing code without extensive reworking, for the sole benefit of making your code more elegant, is simply not something companies are interested in doing.
The likelihood that Broadcom would have any interest in the kind of extensive work it would take to rewrite their own code to incorporate that part of the Linux driver, for no benefit whatsoever to the end user, is very low.
Just how likely is it that there is anything in the Linux driver that would be useful to Broadcom? Broadcom already has fully functional proprietary drivers for their chips.
And is it still you? Or a copy of you running on the artificial brain? If it is a copy, when does it cease to be you?
What makes you think the RIAA has a monopoly?
There are plenty of non-RIAA music providers.
For example, if a state wanted to have a $50/year tax on all automobiles in the state, there would be no problem under Article I, Section 9. They would not have to limit the tax to automobiles that were purchased in the state.