What the hell are you talking about? I am not sure what you are trying to say here, but Verilog and VHDL sure as hell aren't procedural languages, at least not when they are synthesized.
Sure, but now you've defeated the whole point by increasing the server load. Not to mention that it's trivial to write a program to analyze the images. Let's not forget, most spammers are commercial entities. They are ready to spend some time breaking security systems.
Yeah, they are probably using hijacked PCs. I'm just saying that you could ban giveaway addresses. For example, I really doubt any legitimate user will be posting URLs with 'xanax' in them. Of course, the spammers could also get smarter.
Maybe you should set up SpamAssassin to filter forum posts. After all, it does a pretty good job of detecting spammy keywords and such. Sort of like Slashdot's filters.
Another possibility is to put in a probation period. Let's say, if you have been registered for less than a week, you cannot post URLs or use HTML in your forum posts. This is a minimal annoyance to legitmate users, but stops spammers dead in their tracks.
No need to custom program anything. The program can grab 20 or 30 different captchas, figure out which images you are using, and simply have a human mark the kitten ones. This function will be implemented in all the spam software if this technique ever becomes widely used.
Also, I fail to see how a word captcha could be guessable. A 5-letter sequence composed of alphanumeric characters would yield a 1/60466176 chance of guessing it right. That's one in 60 million. You'd be better-off playing the lottery.
Why don't you just implement an auto-ban filter? Attempt to post a URL with 'xanax' or 'casino gambling' in it, and you get your IP permanently banned.
This is even worse than logic puzzles. How many unique kitten pictures is that thing going to have? Ten? Twenty? Maybe fifty? All you have to do to is to get 60% of the kitten pictures programmed into the spambot. Then you just have to compute a CRC of each image served, and bang, you have cracked it. And it's not like it's any better than the scrambled text authentication. If you wanted to reduce server load, you could generate and cache a couple of thousand unique text strings. This approach makes it rather simple for the spambot.
I have seen some captchas that ask users in plain text to solve a simple arithmetic or logic problem.
This is not a good captcha. If someone wants to flood the forums, it takes about 3 minutes to write a regexp to crack these. You aren't going to implement more than 20 or so different logic puzzles, and it's rather trivial to automatically parse these. Also remember that you only need a 5-10% success rate to completely shitflood the forums. I don't think it's possible to create a captcha that is usable by vision-impaired users, except maybe a sound recording (and even that's trivial to run through a voice recognition program).
Actually, almost anything that cannot be implemented as a simple logic circuit can benefit from an operating system. First, multitasking is rampant in many applications. Even the blood pressure monitor needs to respond to user input when it's busy doing something. This is multitasking, although it's usually handled on the application level. However, many applications can benefit from threading support, especially if they need to do stuff like animating a graphical display while doing something else. With a more complicated device, like a system-on-a-chip, an OS becomes really useful because it provides some abstraction from the underlying hardware. This is important -- you want to be able to switch to a cheaper chip from a differnt vendor without rewriting a ton of code. Not to mention, the chip vendor often writes drivers for its hardware, so it takes less work to implement stuff. Of course, if you need networking support, an OS is pretty much a given. In short, embedded operating systems are becoming increasingly relevant.
Holy shit, you do a lot of PIC development? That's amazing dude. OBVIOUSLY you don't need an OS to run an 8-bit microcontroller with 128 bits of RAM. In fact, you don't even have a choice there. However, even microwaves are starting to appear with color LCD displays and friendlier user interfaces. Those obviously need an OS. In fact, the vast majority of products out there use 16 or 32-bit microcontrollers with lots of memory and ROM. They most certainly require an operating system. The market for 8-bit applications is becoming rather marginal these days, they are used mostly for support functions, if at all.
Your post doesn't make much sense. You have more than 60GB of music, but you can fit it on 5 minidiscs? That's ignoring the fact that your collection would be extremely easy to manage through iTunes, that the iPod interface is designed to handle hundreds of artists, and that you aren't going to destroy a hard drive with any reasonable activity (short of dropping it on concrete from 6 feet). The iPod has like a 32 MB RAM buffer, so it only spins up the hard drive once every 15 minutes or so.
It's painfully obvious that the author of the article is still stuck in the 90s. Of course, most people that haven't owned an iPod also think this way. The main thing with an iPod (or any HDD-based music player) is that you have _all_ your music on it. You are not limited to the songs on a particular disc, and you can find any song in your collection in under 20 seconds. Not to mention, this is all on one compact device. I guess if I wanted to look like a dork and carry around 30 1GB minidiscs, swap them every 5 minutes, and deal with the hassle of remembering which music is on which disc, I would go with that format. Not to mention that at Sony prices, a player and 30 minidiscs would probably run you a lot more than $300. But hey, you get to stand out from the crowd by being the guy with a dorky player.
With the patent situation these days, you never know. Everyone thought JPEG was patent-free, until someone produced a patent. Most patent holders wait until something is being widely used before they start demanding money. Nobody is going to sue some tiny company that produces Vorbis-enabled MP3 players -- there is no money to be had there. Apple would be a much juicier target.
Well, it's rather difficult to quantify the quality of programmers. Some programmers can generate large quantities of relatively bug-free code in short periods of time. Other programmers are good at coming up with good algorithms and solving complicated problems. Still other programmers are not as good at either, but can be excellent software engineers / system architects. There is quite a number of different skills that may be needed in different circumstances.
Ogg Vorbis is nice, but the legal liability for Apple would be too high to implement it. Vorbis most likely violates a large number of patents, and a company the size of Apple is a rather attractive target. For them, it's much cheaper to pay for MP3 and AAC than to risk getting sued. Not to mention, if the iPod can support more audio formats than it does already, they are not using the hardware efficiently. I doubt Vorbis could be added cheaply or easily.
You are reading the wrong books. Look at some of the IEEE publications, they are full of cutting-edge research results that nobody had a chance to apply yet.
Fluorescent lights do not flicker at 60Hz. If you are in a really old building with 70s era lighting, they will flicker at 120Hz. Any decent lighting fixture made in the 90s will have a switching power supply for driving the tube; those do not exhibit flicker, since the tube runs off of DC.
What I think is the real culprit is all the acoustic crap put out by switching power supplies. Many of the old designs run well within the audible frequency range, and the high-pitched squeals can annoy people even if they can't hear them very well. Most power supplies used in consumer equipment are guilty of this sin, since they use bargain-bin components that cannot run at high frequencies. This may be the real cause behind the electrosenstivity bullshit.
The main switching power supply in the monitor often runs at a frequency less than 20kHz, since most monitors are cheap, 20 year old designs. This is considered unacceptable for contemporary switching power supplies (the ones we design at work are like 500kHz), but they are still used in consumer equipment.
Most people feel more comfortable with an LCD, since they don't flicker and have much sharper text. I can look at an LCD all day, but a CRT makes my eyes tired after a few hours. Computer labs are especially problematic, since monitors tend to be set for 60Hz and you get all that flicker in your peripheral vision.
Any observational study is subject to flaws, due to confounding factors. Property along high voltage power lines tends to be significantly cheaper, since these power lines are usually built in lightly populated or cheap areas. Most rich neighborhoods have underground cables, while most poor ones have overhead lines. Any statistics you are going to get from comparing houses are going to be dominated by these effects, rendering any correlation meaningless unless these factors are carefully eliminated (which is almost impossible to do).
The horizontal output stage in a TV makes noise at around 16kHz. You aren't experiencing EMI, you are just hearing a high-pitched noise. Many switching power supplies produce high-pitched whines and noises. I remember hearing those when I was 5 or so. By the time you're 25, your hearing gets much worse, and most people cannot hear them anymore.
Hahhahahaha... shielded by the metal tank? Bullshit. Nothing can shield EMI from an automotive ignition coil when it's up close like that. About the only thing that can be done is reducing the dv/dt by using resistive spark plug wires, and you can still detect EMI from an ignition system a mile away.
I also have a hard time believing the wifi claim. A wifi access point has less power than a single frigging cellphone. Can you detect when someone's cellphone is turned on? Maybe you should make a double-blind experiment like that. Get 2 friends, have one put a cellphone inside a box (either on or off) and have the other one present you with the box. If you can reliably detect when the cellphone is on, I'm sure a lot of researchers would be interested in working with you.
Slashdot Mirror
What the hell are you talking about? I am not sure what you are trying to say here, but Verilog and VHDL sure as hell aren't procedural languages, at least not when they are synthesized.
Sure, but now you've defeated the whole point by increasing the server load. Not to mention that it's trivial to write a program to analyze the images. Let's not forget, most spammers are commercial entities. They are ready to spend some time breaking security systems.
Have you heard about Verilog, VHDL, and FPGAs? Go google those sometime. Desktop chip manufacturing... ROFL
Yeah, they are probably using hijacked PCs. I'm just saying that you could ban giveaway addresses. For example, I really doubt any legitimate user will be posting URLs with 'xanax' in them. Of course, the spammers could also get smarter.
Maybe you should set up SpamAssassin to filter forum posts. After all, it does a pretty good job of detecting spammy keywords and such. Sort of like Slashdot's filters.
Another possibility is to put in a probation period. Let's say, if you have been registered for less than a week, you cannot post URLs or use HTML in your forum posts. This is a minimal annoyance to legitmate users, but stops spammers dead in their tracks.
No need to custom program anything. The program can grab 20 or 30 different captchas, figure out which images you are using, and simply have a human mark the kitten ones. This function will be implemented in all the spam software if this technique ever becomes widely used.
Also, I fail to see how a word captcha could be guessable. A 5-letter sequence composed of alphanumeric characters would yield a 1/60466176 chance of guessing it right. That's one in 60 million. You'd be better-off playing the lottery.
Why don't you just implement an auto-ban filter? Attempt to post a URL with 'xanax' or 'casino gambling' in it, and you get your IP permanently banned.
This is even worse than logic puzzles. How many unique kitten pictures is that thing going to have? Ten? Twenty? Maybe fifty? All you have to do to is to get 60% of the kitten pictures programmed into the spambot. Then you just have to compute a CRC of each image served, and bang, you have cracked it. And it's not like it's any better than the scrambled text authentication. If you wanted to reduce server load, you could generate and cache a couple of thousand unique text strings. This approach makes it rather simple for the spambot.
I have seen some captchas that ask users in plain text to solve a simple arithmetic or logic problem.
This is not a good captcha. If someone wants to flood the forums, it takes about 3 minutes to write a regexp to crack these. You aren't going to implement more than 20 or so different logic puzzles, and it's rather trivial to automatically parse these. Also remember that you only need a 5-10% success rate to completely shitflood the forums. I don't think it's possible to create a captcha that is usable by vision-impaired users, except maybe a sound recording (and even that's trivial to run through a voice recognition program).
Bullshit. Companies do it all the time, and that defense never held water in patent cases.
You don't have jack shit if it's a small company. And if you have stock in the company, you get all that without any of the work.
Actually, almost anything that cannot be implemented as a simple logic circuit can benefit from an operating system. First, multitasking is rampant in many applications. Even the blood pressure monitor needs to respond to user input when it's busy doing something. This is multitasking, although it's usually handled on the application level. However, many applications can benefit from threading support, especially if they need to do stuff like animating a graphical display while doing something else. With a more complicated device, like a system-on-a-chip, an OS becomes really useful because it provides some abstraction from the underlying hardware. This is important -- you want to be able to switch to a cheaper chip from a differnt vendor without rewriting a ton of code. Not to mention, the chip vendor often writes drivers for its hardware, so it takes less work to implement stuff. Of course, if you need networking support, an OS is pretty much a given. In short, embedded operating systems are becoming increasingly relevant.
Holy shit, you do a lot of PIC development? That's amazing dude. OBVIOUSLY you don't need an OS to run an 8-bit microcontroller with 128 bits of RAM. In fact, you don't even have a choice there. However, even microwaves are starting to appear with color LCD displays and friendlier user interfaces. Those obviously need an OS. In fact, the vast majority of products out there use 16 or 32-bit microcontrollers with lots of memory and ROM. They most certainly require an operating system. The market for 8-bit applications is becoming rather marginal these days, they are used mostly for support functions, if at all.
Your post doesn't make much sense. You have more than 60GB of music, but you can fit it on 5 minidiscs? That's ignoring the fact that your collection would be extremely easy to manage through iTunes, that the iPod interface is designed to handle hundreds of artists, and that you aren't going to destroy a hard drive with any reasonable activity (short of dropping it on concrete from 6 feet). The iPod has like a 32 MB RAM buffer, so it only spins up the hard drive once every 15 minutes or so.
It's painfully obvious that the author of the article is still stuck in the 90s. Of course, most people that haven't owned an iPod also think this way. The main thing with an iPod (or any HDD-based music player) is that you have _all_ your music on it. You are not limited to the songs on a particular disc, and you can find any song in your collection in under 20 seconds. Not to mention, this is all on one compact device. I guess if I wanted to look like a dork and carry around 30 1GB minidiscs, swap them every 5 minutes, and deal with the hassle of remembering which music is on which disc, I would go with that format. Not to mention that at Sony prices, a player and 30 minidiscs would probably run you a lot more than $300. But hey, you get to stand out from the crowd by being the guy with a dorky player.
With the patent situation these days, you never know. Everyone thought JPEG was patent-free, until someone produced a patent. Most patent holders wait until something is being widely used before they start demanding money. Nobody is going to sue some tiny company that produces Vorbis-enabled MP3 players -- there is no money to be had there. Apple would be a much juicier target.
Well, it's rather difficult to quantify the quality of programmers. Some programmers can generate large quantities of relatively bug-free code in short periods of time. Other programmers are good at coming up with good algorithms and solving complicated problems. Still other programmers are not as good at either, but can be excellent software engineers / system architects. There is quite a number of different skills that may be needed in different circumstances.
Where I went to school, Diff EQ wasn't even required for CS majors. And it wasn't even that hard.
Ogg Vorbis is nice, but the legal liability for Apple would be too high to implement it. Vorbis most likely violates a large number of patents, and a company the size of Apple is a rather attractive target. For them, it's much cheaper to pay for MP3 and AAC than to risk getting sued. Not to mention, if the iPod can support more audio formats than it does already, they are not using the hardware efficiently. I doubt Vorbis could be added cheaply or easily.
You are reading the wrong books. Look at some of the IEEE publications, they are full of cutting-edge research results that nobody had a chance to apply yet.
Fluorescent lights do not flicker at 60Hz. If you are in a really old building with 70s era lighting, they will flicker at 120Hz. Any decent lighting fixture made in the 90s will have a switching power supply for driving the tube; those do not exhibit flicker, since the tube runs off of DC.
What I think is the real culprit is all the acoustic crap put out by switching power supplies. Many of the old designs run well within the audible frequency range, and the high-pitched squeals can annoy people even if they can't hear them very well. Most power supplies used in consumer equipment are guilty of this sin, since they use bargain-bin components that cannot run at high frequencies. This may be the real cause behind the electrosenstivity bullshit.
The main switching power supply in the monitor often runs at a frequency less than 20kHz, since most monitors are cheap, 20 year old designs. This is considered unacceptable for contemporary switching power supplies (the ones we design at work are like 500kHz), but they are still used in consumer equipment.
Most people feel more comfortable with an LCD, since they don't flicker and have much sharper text. I can look at an LCD all day, but a CRT makes my eyes tired after a few hours. Computer labs are especially problematic, since monitors tend to be set for 60Hz and you get all that flicker in your peripheral vision.
Any observational study is subject to flaws, due to confounding factors. Property along high voltage power lines tends to be significantly cheaper, since these power lines are usually built in lightly populated or cheap areas. Most rich neighborhoods have underground cables, while most poor ones have overhead lines. Any statistics you are going to get from comparing houses are going to be dominated by these effects, rendering any correlation meaningless unless these factors are carefully eliminated (which is almost impossible to do).
The horizontal output stage in a TV makes noise at around 16kHz. You aren't experiencing EMI, you are just hearing a high-pitched noise. Many switching power supplies produce high-pitched whines and noises. I remember hearing those when I was 5 or so. By the time you're 25, your hearing gets much worse, and most people cannot hear them anymore.
Hahhahahaha... shielded by the metal tank? Bullshit. Nothing can shield EMI from an automotive ignition coil when it's up close like that. About the only thing that can be done is reducing the dv/dt by using resistive spark plug wires, and you can still detect EMI from an ignition system a mile away.
I also have a hard time believing the wifi claim. A wifi access point has less power than a single frigging cellphone. Can you detect when someone's cellphone is turned on? Maybe you should make a double-blind experiment like that. Get 2 friends, have one put a cellphone inside a box (either on or off) and have the other one present you with the box. If you can reliably detect when the cellphone is on, I'm sure a lot of researchers would be interested in working with you.