Apparently health insurance costs vary drastically with location. It's no wonder people are so polarized about the urgency of health care reform. If you live in an area that already has affordable health insurance, chances are you think little needs to be done. If you live in an area that does not have affordable health insurance, chances are you think much needs to be done.
For instance, $120/month in Maricopa County, AZ would get me a PPO plan with a $2500 deductible and 0% coinsurance. There are dozens of additional choices with prices in the same ballpark. Some of them look downright reasonable. $150/month in Rensselaer County, NY would get me a hospital-only plan; nothing else would be covered. (A PPO plan would cost at least $500/month, and many services such as office visits still wouldn't be covered at that price.)
Maybe C and C++ aren't common for introductory programming courses anymore, but they used to be for a time. About 10 years ago:
I took an introductory C programming "class" as part of the Columbia University Science Honors Program for high school students. It was the only programming course offered. The Computer Science Advanced Placement exam switched from Pascal to C++. At Rensselaer Polytechnic Institute, Computer Science I and II were taught in C++.
What I want to know is, why was a few hundred thousand reCAPTCHA challenges enough to have a reasonable chance of getting a duplicate? Shouldn't the number of possible challenges be several orders of magnitude higher to discourage this kind of attack?
The only reason what you quoted doesn't work is because in the case of a tie (50% of the time), it is using the initial ordering. That is the source of bias in your example, and it is avoidable.
Some "reasonable sounding" methods don't actually work - e.g. attach random numbers to each list item and sort the list by these numbers (1).
That method does work as long as you guarantee the uniqueness of the random numbers--in other words, a tie shouldn't mean leaving elements in place, but rather repeating the algorithm on each subset of the list where the random numbers came up equal. I see no advantage to it over the Fisher-Yates shuffle (which should be the default solution to the problem), but it can be made to work.
$75/month? $508.04/month is the cheapest plan listed for me, other than a $158.41/month hospital-only plan. I'm a healthy non-smoking male in my late 20's. The insurance companies have failed us.
I write software for a living. Yet because I have a bachelor's degree in computer science and mathematics, I am apparently as qualified as some of those people to make a scientific claim regarding global warming: 935 of the people on the list have degrees in computer science and/or math.
One of my relatives is an athletic trainer with a masters degree in sports medicine. She is equally qualified: 2,327 of the people on the list have degrees in medicine.
Another relative has worked in sales and management, with a bachelor's degree in industrial engineering. He is equally qualified: 7,281 of the people on the list have degrees in unspecified engineering disciplines.
Come to think of it, almost every living adult in my family--parents, siblings, aunts, uncles, cousins, and grandparents--meets the single criterion (aside from agreeing with the petition statement) to be included. And none of them are scientists, or have any sort of specialized knowledge of climate science.
An appeal to false authority is not a way to make an effective argument.
I present here a way of computing the nth decimal digit of pi (or any other base) by using more time than the [BBP] algorithm but still with very little memory.
At the risk of a "whoosh!", in context HOPO means "hammer-on/pull-off". When playing a Guitar Hero style game, ordinary notes are played by holding a colored fret button and pushing the strum bar. Some fast notes can be designated as "HOPO" notes, only requiring pushing the appropriate fret button if the previous note was played correctly. This is a simulation of the "hammer on" and "pull-off" methods of playing notes on a real guitar without strumming.
Difficult tracks may have long sequences of "HOPO" notes. If you miss one, it is important to strum again quickly or else the rest of the sequence of notes will be counted as misses. Latency delays notification of a missed note.
They can't run the engine out in advance of your input
Slightly offtopic, but that's exactly the solution I came up with when idly trying to think of a way for old multiplayer console games to be playable over the internet. (This assumes local graphics rendering and vastly superior hardware to the original console, which is not the case with the OnLive service.) Prediction can be done even when the core game engine wasn't designed for it, if the game is running inside a meta-engine like an emulator.
Assuming a coarse input sample rate, and that relatively little computational power is needed to maintain game state (enough to be able to reconstruct graphics/sound at a given time but without actually doing so), then simply make a copy of the game state each time a non-local player's input is checked. Optionally, with any spare CPU time calculate alternate branches of game states--"what if player 2 had pressed X here?" If possible, merge any branches that result in identical game states--if player 2 pressing X for any 30 ms block in the last 100 ms results in the same outcome, for instance. Give preference to game states where input remains unchanged: if working on a "what if player 2 had pressed X" branch, then first calculate later game states where player 2 is still holding down X. Unlikely input scenarios that involve, for example, the player pressing a new button every few ms should be considered last.
All input should be sent to remote players with a timestamp relative to the game reset time. When remote player input arrives, switch to the game state branch corresponding to that input and timestamp, discard any game states descending from any other input at that time, and recalculate the present game state from there (if it hasn't been calculated already). Note the input latency, and use that to determine which game state branches should be precalculated first--if input tends to come in 80-100 ms late, then focus on precalculating branches that diverge 80-100 ms behind the present. Exchange latency information so that games can be approximately synchronized between players.
If each player controls one on-screen character, then the player's own character should be highly responsive even if the other players' characters stutter or skip around on screen a little. With at least some types of games this should be acceptable.
I'm sure you can provide even one example of where an emulated machine has ever had a vulnerability allowing a program running in the emulated hardware to do something malicious to the (usually) unrelated host hardware.
VICE is a program that runs on a Unix, MS-DOS, Win32, OS/2, Acorn RISC OS or BeOS machine and executes programs intended for the old 8-bit Commodore computers. The current version emulates the C64, the C128, the VIC20, all the PET models (except the SuperPET 9000, which is out of line anyway), the PLUS4 and the CBM-II (aka C610).
There is a format string vulnerability in the handling of the monitor "memory dump" command. If the string to be output contains any % sign, it is interpreted as a command for the output, normally resulting in a crash. Even more sophisticated exploits, like arbitrary code execution on the host machine, are possible.
Now, the obvious counter-argument is that such a vulnerability is possible in a wide variety of programs, not just emulators. One widely known game example (and games are obviously allowed on the iPhone): http://wiibrew.org/wiki/Twilight_Hack
The Twilight Hack works by employing a lengthy character name for the horse in the game ('Epona') in order to facilitate a stack smash. This gets triggered when talking to the man next to you when you start the saved game as he loads the name to use it in his dialog or upon attempting to enter the next zone, before the man talks to you and reminds you to go the other way to get the horse.
At some point, most people would have had a "huh... what happened to that ticket" thing pop up in their brains, prompting action.
That happened to me with a speeding ticket. About a month after mailing the ticket in with a not guilty plea, I started to worry that I had thrown out a court date notification (perhaps stuffed inside junk mail shaped like a newspaper), or that the ticket I mailed never arrived. I called the court, and a clerk told me that everything was in order, and to expect a court date months later. She was right--I eventually received a court date some 8 months later.
I see Latin scientific names as a separate namespace used to avoid ambiguity between descriptions and classification/labels. That way two things can appear alike (resulting in similar descriptions) but can still be concisely distinguished based on less obvious characteristics.
While you raise some good points, I think you overestimate the need to be computer savvy to play MMO games. I've met a number of people that use their computers almost exclusively to play EverQuest--I had to walk one through the process of copying and pasting text in Notepad, for example.
If you're going to do that, why not create a genuine random dot stereogram, and encode the data in the "random" dot pattern of the first column? You can probably fit quite a bit of data in there compared to ordinary steganographic techniques, since it's expected that there will be a lot of random-looking noise in such a picture.
Not necessarily. Ever tried to uninstall Outlook Express from Windows XP without following specific directions? http://www.tweakxp.com/article37318.aspx If you don't follow them, and just delete the files, they reappear after you reboot (if I remember right).
The problem here is that some people sign up on a site that exists to share personal information, and their friends run apps that give away personal information and tell them they're doing it, and are then surprised.
THAT is the problem as I understand it: apparently you can't deny information to apps that your friends have authorized but you have not.
As of a few months ago, kopete occasionally dropped messages silently (confirmed via other channels). I switched to pidgin and no longer had to restart the program each time a "still there?" question went unanswered.
For webcam support on yahoo, gyachi works nearly flawlessly for me.
This same teacher graded tests on a type of curve so if the high score was 47 out of 50 then that person got 100% and all other scores were X out of 47. He would also remind everyone at the beginning of the test that if no one answers any questions everyone would get 100%
Shouldn't everyone's grades be undefined if that happens?
Slashdot Mirror
Apparently health insurance costs vary drastically with location. It's no wonder people are so polarized about the urgency of health care reform. If you live in an area that already has affordable health insurance, chances are you think little needs to be done. If you live in an area that does not have affordable health insurance, chances are you think much needs to be done.
For instance, $120/month in Maricopa County, AZ would get me a PPO plan with a $2500 deductible and 0% coinsurance. There are dozens of additional choices with prices in the same ballpark. Some of them look downright reasonable.
$150/month in Rensselaer County, NY would get me a hospital-only plan; nothing else would be covered. (A PPO plan would cost at least $500/month, and many services such as office visits still wouldn't be covered at that price.)
Maybe C and C++ aren't common for introductory programming courses anymore, but they used to be for a time. About 10 years ago:
I took an introductory C programming "class" as part of the Columbia University Science Honors Program for high school students. It was the only programming course offered.
The Computer Science Advanced Placement exam switched from Pascal to C++.
At Rensselaer Polytechnic Institute, Computer Science I and II were taught in C++.
What I want to know is, why was a few hundred thousand reCAPTCHA challenges enough to have a reasonable chance of getting a duplicate? Shouldn't the number of possible challenges be several orders of magnitude higher to discourage this kind of attack?
The only reason what you quoted doesn't work is because in the case of a tie (50% of the time), it is using the initial ordering. That is the source of bias in your example, and it is avoidable.
That method does work as long as you guarantee the uniqueness of the random numbers--in other words, a tie shouldn't mean leaving elements in place, but rather repeating the algorithm on each subset of the list where the random numbers came up equal.
I see no advantage to it over the Fisher-Yates shuffle (which should be the default solution to the problem), but it can be made to work.
$75/month? $508.04/month is the cheapest plan listed for me, other than a $158.41/month hospital-only plan. I'm a healthy non-smoking male in my late 20's.
The insurance companies have failed us.
Will we? Images of Pluto are still pretty underwhelming. It may be small and poorly lit, but it's a heck of a lot closer to us than any exoplanet.
http://www.petitionproject.org/qualifications_of_signers.php
Note that the only requirement regarding background is a degree (bachelor's or higher) in a "related" scientific field.
I write software for a living. Yet because I have a bachelor's degree in computer science and mathematics, I am apparently as qualified as some of those people to make a scientific claim regarding global warming: 935 of the people on the list have degrees in computer science and/or math.
One of my relatives is an athletic trainer with a masters degree in sports medicine. She is equally qualified: 2,327 of the people on the list have degrees in medicine.
Another relative has worked in sales and management, with a bachelor's degree in industrial engineering. He is equally qualified: 7,281 of the people on the list have degrees in unspecified engineering disciplines.
Come to think of it, almost every living adult in my family--parents, siblings, aunts, uncles, cousins, and grandparents--meets the single criterion (aside from agreeing with the petition statement) to be included. And none of them are scientists, or have any sort of specialized knowledge of climate science.
An appeal to false authority is not a way to make an effective argument.
What about this?
We already have the Legend of Neil (contains strong language)...
At the risk of a "whoosh!", in context HOPO means "hammer-on/pull-off". When playing a Guitar Hero style game, ordinary notes are played by holding a colored fret button and pushing the strum bar. Some fast notes can be designated as "HOPO" notes, only requiring pushing the appropriate fret button if the previous note was played correctly. This is a simulation of the "hammer on" and "pull-off" methods of playing notes on a real guitar without strumming.
Difficult tracks may have long sequences of "HOPO" notes. If you miss one, it is important to strum again quickly or else the rest of the sequence of notes will be counted as misses. Latency delays notification of a missed note.
Slightly offtopic, but that's exactly the solution I came up with when idly trying to think of a way for old multiplayer console games to be playable over the internet. (This assumes local graphics rendering and vastly superior hardware to the original console, which is not the case with the OnLive service.) Prediction can be done even when the core game engine wasn't designed for it, if the game is running inside a meta-engine like an emulator.
Assuming a coarse input sample rate, and that relatively little computational power is needed to maintain game state (enough to be able to reconstruct graphics/sound at a given time but without actually doing so), then simply make a copy of the game state each time a non-local player's input is checked. Optionally, with any spare CPU time calculate alternate branches of game states--"what if player 2 had pressed X here?" If possible, merge any branches that result in identical game states--if player 2 pressing X for any 30 ms block in the last 100 ms results in the same outcome, for instance. Give preference to game states where input remains unchanged: if working on a "what if player 2 had pressed X" branch, then first calculate later game states where player 2 is still holding down X. Unlikely input scenarios that involve, for example, the player pressing a new button every few ms should be considered last.
All input should be sent to remote players with a timestamp relative to the game reset time. When remote player input arrives, switch to the game state branch corresponding to that input and timestamp, discard any game states descending from any other input at that time, and recalculate the present game state from there (if it hasn't been calculated already). Note the input latency, and use that to determine which game state branches should be precalculated first--if input tends to come in 80-100 ms late, then focus on precalculating branches that diverge 80-100 ms behind the present. Exchange latency information so that games can be approximately synchronized between players.
If each player controls one on-screen character, then the player's own character should be highly responsive even if the other players' characters stutter or skip around on screen a little. With at least some types of games this should be acceptable.
Will this do as an example?
http://seclists.org/bugtraq/2004/Jun/211
Now, the obvious counter-argument is that such a vulnerability is possible in a wide variety of programs, not just emulators. One widely known game example (and games are obviously allowed on the iPhone): http://wiibrew.org/wiki/Twilight_Hack
That happened to me with a speeding ticket. About a month after mailing the ticket in with a not guilty plea, I started to worry that I had thrown out a court date notification (perhaps stuffed inside junk mail shaped like a newspaper), or that the ticket I mailed never arrived. I called the court, and a clerk told me that everything was in order, and to expect a court date months later. She was right--I eventually received a court date some 8 months later.
I think what he meant is that on average adjacent primes get further apart (see prime number theorem).
I see Latin scientific names as a separate namespace used to avoid ambiguity between descriptions and classification/labels. That way two things can appear alike (resulting in similar descriptions) but can still be concisely distinguished based on less obvious characteristics.
While you raise some good points, I think you overestimate the need to be computer savvy to play MMO games. I've met a number of people that use their computers almost exclusively to play EverQuest--I had to walk one through the process of copying and pasting text in Notepad, for example.
If you're going to do that, why not create a genuine random dot stereogram, and encode the data in the "random" dot pattern of the first column? You can probably fit quite a bit of data in there compared to ordinary steganographic techniques, since it's expected that there will be a lot of random-looking noise in such a picture.
There was a Slashdot article when the contest started, too:
http://yro.slashdot.org/story/09/08/19/1626217/Wired-Writer-Disappears-Find-Him-and-Make-5k
Not necessarily. Ever tried to uninstall Outlook Express from Windows XP without following specific directions?
http://www.tweakxp.com/article37318.aspx
If you don't follow them, and just delete the files, they reappear after you reboot (if I remember right).
Ah, I stand educated. Thank you.
If that page works as advertised, it needs to be displayed more prominently here. (Mod parent up?)
Not quite--
THAT is the problem as I understand it: apparently you can't deny information to apps that your friends have authorized but you have not.
As of a few months ago, kopete occasionally dropped messages silently (confirmed via other channels). I switched to pidgin and no longer had to restart the program each time a "still there?" question went unanswered.
For webcam support on yahoo, gyachi works nearly flawlessly for me.
Shouldn't everyone's grades be undefined if that happens?
Is buying a $400 Visa/Mastercard gift card, then using that to shop online, an option?