He wasn't trying to make it look like he was downloading material. To continue your analogy, making a substance that smells and looks like pot would be sending false data to the torrent. He didn't do that. His program is observational in nature. It's a bit like calling the cops in the guy with the camera in the back alley because you think its suspicious. It doesn't transfer anything, it doesn't accept anything from seeds. Even if you tried sending garbage, you'd be caught by the md5 hashes and blocked / banned.
If your goal is to examine the nature of torrents and how different goals result in different patterns, a tool like this is vital. It's not at all straightforward that it's unethical or illegal to connect to the swarm to gather user statistics.
I've heard that disney bought up florida land under a secretive context, and it was quite a bruhaha at the time when it was discovered that they were building a resort there. Perhaps google is worried that if word gets out they're making a large investment there, that someone will start looking for where google would want to locate (perhaps a large unused building located in a large tech heavy town, say an old AOL building?)
Page rank was/is hands above its peers at introduction, but it's not pagerank that makes google spectacular. It's about making pagerank fast. Google successfully improved web search, and then made it scalable both in number of users and in pages indexed.
Open source is good at providing new ideas, but high performance doesn't have a huge market for users. Many of Google's most important optimizations require a special configuration and hardware, like thousands of computers laden with RAM, power etc. As it happens, wikipedia often times cannot handle it's users demands. Their search feature is sometimes handed off to google. Lets also not forget the amount of data required to be stored. Sure, you can download tons of websites, but you can't discard everything; you'll need the url, relevant links, and context, excerpts to show users, etc.
Of course, all this pessmism hasn't stopped people from trying. If you discover a list of sites running Nutch, you'll notice most are niche sites, either offering to search a single website, or subtopic. The only general web search deployment I found was fast but crappy.
This comment hurts my brain. You bitch about TLAs but they're not even acronyms. Do you even know what TLA stands for?!
Lockless means it's doing something without locking everyone else from the data. Sometimes this means optimistic resolution (everyone try, and if it looks like it screwed up, try again!), sometimes it means keeping local read copies, sometimes it means something new and/or crazy. Lockless approaches are used when you have data that lots of threads must share but efficiency concerns or non-blocking requirements force you away from simply using a lock and blocking when someone else is playing with the data.
A radix-tree, as opposed to "radix-free", is a data structure used in certain applications, with operations dependent on the length of the key rather than the amount of data stored. In 2.6.20 (and others), it's used to organize some information about the page cache.
This code is associated with the RCU, which you may recall is part of an SCO lawsuit. If you're interested in any other feature or changes, the kernel newbies site is instrumental!
I haven't shopped at ebay in a long time, but apparently one can withdraw a bid. So it should come as no surprise that scammers use this feature to eliminate the risk of overbidding. Meanwhile, the consumer is competing with demand that doesn't actually exist. Apparently it's popular enough that ebay offers a Second Chance feature, that lets a seller offer the second highest bidder the option to buy at their highest losing bid should the winner fail to complete the transaction.
I guess as long as ebay is interested in more sales and higher prices, there's too much incentive to address the problem poorly.
"Oh, the horrors! I've had this happen to me many times, as well, especially on Linux. Not even a kill -9 would get rid of the wedged process. Why is there even such a thing as "non-interruptible sleep"? If I don't need the process anymore, I should be able to get rid of it no matter what."
I haven't yet read the internals of the linux kernel, but I understand that disk dma requests are a major source of non-interruptible sleep. I imagine it has to do with locking the RAM the disk is going to write to safe from other processes. But from my studies of MINIX, I recall that the DMA controllers were a bit limited, so its possible DMA in linux is done to some buffer and then copied into userspace. This would make my whole argument invalid =(
Fantasy land material. Wal-mart's data centers would eat MS products alive. Recall that every transaction is being logged there. About seven years ago, my university recieved a donation of one of their district processing mainframes: something like an 82 way pentium 2 setup. Fantastic sounding stuff, but it was a) too slow for their (regional) needs, and b) too damned hard to make fast (NUMA).
If Walmart was dissatisfied with Linux, somehow I think Windows would be their last pick. Which makes me wonder, what are they using now? Linux? Solaris?
As far as I know, the GMA 9xx series is a couple generations behind, performance wise. It should play quake3 and UT2k4 just fine, but it seems to have trouble with the Doom 3 engine, and I suspect the new UT engine will also be unplayable. On the windows side, it doesn't work with halflife2 either. Seems the most likely kind of game to fail is a new FPS. But I hear aero and Xgl/AIGLX work fine, so you may be satisfied with the current Intel offerings. The wikipedia page seems like a good place to start researching if you're still interested.
Why does liscencing a patent require closed source drivers? If I recall correctly, the patent system requires disclosure as reciprocation for limited monopolies. I believe there are open source 3d drivers for the savage3 chipset, which is strange given the circumstances claimed.
Sadly, while your energy production peaks coincide with peak demand, you're still paid a flat rate for the production. I hear that power companies charge some high energy users more during peak hours, and less during the night / troughs of low demand. Paying out more for energy during peak hours would probably improve the incentives to go solar, but I guess the cost of monitoring this would be huge compared to reading a number every month.
The news here isn't that they're using x86 architecture. It's that they're using Intel chips. They already have AMD Opterons for sale, and they're adding Xeons. And Solaris 10 does run on x86. Some see it as a concession that their server CPU designs are little more than a niche market. That diagnosis is probably correct, and if Sun wants to ever dig themselves out of the "Sun is dying" meme, they'll have to start taking advantage of the fact that their competitors are engaged in a price war, one that's also cutting into profits. Sun can still pride themselves on quality server hardware, support contracts and integration, even if they aren't the designer behind most their chips.
Meanwhile, their Niagra still has some niche applications, and will grow as software is designed for dual and quad core chips. If Niagra does what they say it will, people will be forced to consider one Niagra unit versus 6 Xeon servers. Xeons may have fallen in price recently, but they're still not cheap, so that's a calculus that Sun might win some day.
They're getting into the business because they have tons of money to invest, they already know the system well enough, and the only significant competition is PayPal. They've been offering "pay services" since about 2000-ish. Remember Adwords? The whole thing's automated and doesn't rely on paypal or other services that compete with google checkout. Way older than charging for Picasa / Google Earth.
They offered the money for one simple reason: adoption rates. I can't even begin to describe the amount of money paypal threw at the altar of adoption rates. By giving money away, they're quickly getting merchants who need an edge over their competitors, or are just worried about their competitors doing the same. And customers have the usual incentives to take up the deal. The reason adoption rates matter is that the value of their service is based on the number of possible transactions that an individual can participate in. The more merchants using it, the more likely you'll want to sign up for a google checkout account, even without the discount offers. And the more consumers with accounts, the more sense it makes to put forth the effort to integrate google checkout into your website. It's a virtuous cycle, but Google needed to give it a push start. I'm not sure one needs to resort to a data mining conspiracy to justify their actions here.
"Linus may fix an issue with the VM subsystem, but usually won't be the person to fix a problem with ext2"
It's kinda interesting that you bring this up, becuase he recently went through and fixed a data loss problem caused by interaction between VM and ext2. Just because someone else is the maintainer of it doesn't mean he can't go through and debug it like anyone else. But I think that there are plenty of people who would like to go through the nvidia source code. They may not all be experts, but I have a sneaking suspcion that the nvidia team isn't either. At first, you'll probably see a few code critiques, during which the module maintainers may be exposed to "kernel best practices" rants. But for all their internal knowledge, there's likely enough documentation in the drivers alone to more than double the efforts at improving nvidia drivers after 6 to 12 months. For example, the nouveau project is already making progress towards an open source driver for nvidia 3d chipsets, without anything more than the binary drivers, cards, and helpful users. Even if a tragically small percentage of OSS developers are able to help nouveau, it's still more than nvidia has dedicated to it.
"Judge Wells asked IBM to help SCO out in any way he could."
Maybe this is why I'll never go to law school, but I'd probably respond to that at that point by placing a loaded gun on the table and tell them "Kill yourself. Now."
On the contrary, when taking the article at face value, I think that the whole intent is to make it *easy* for Microsoft to implement the first time, because it's already done. This is of course backwards from how a community standard should work, it should be an effort that is repeatable. Instead we have whatever crap their contractors turned in, with apparent flaws turned into requirements. I doubt even Microsoft could write a second compatible handler for this document format. Well, perhaps instead of "on the contrary" I should say "to clarify"; the standard appears to be designed so that any implementation but the original is near impossible.
It sounds like pretty much like business as usual for MSFT, although describing in 6,000 pages how hard it would be to create an interoperating product is new. Their format is the standard, even the flaws that they didn't fix before release.
The engineering profession works mostly the same way in the US. The Texas deal was them trying to jumpstart a Software Engineering field. If you look at the specializations, there's a huge preference for building codes. Fire safety engineering. Civil Engineering. There's no computer chip engineering specialty. Basically, the definition of "engineer" to them is someone who takes thermodynamics.
But if you recall that engineering comes from designing new and useful things, rather than specifically building a bridge or an engine, the case for Software Professional Engineers gets strong. There are tons of public systems in use today that rely on software, but the state currently has no means of assessing quality. The reason the US hasn't had to re-examine this in great detail is the exception clause: if you're work is for a specific employer "in industry" then the company takes liability, not you. Of course, most software comes with "NO GUARENTEE FOR ANY PARTICULAR USE" etc. Somehow, the DoD manages to survive simply by assessing bidders, although I hear the Grand Challenge was partly started as an alternative to the failures DARPA had in getting products of value from industry.
It sounds like there is a valid case to be had for professional licensing of software developers. Whether that term is "engineer" is moot; the end result would likely look and operate in the exact same way. The strongest counter-argument, and the one I never hear from people invested in the Professional Engineer licensing program, is that software is too young to encode a lasting best practices and design techniques. Five years ago, you wouldn't likely have heard of unit testing. Verification and validation of software was something PhDs did; now I know of a few defense contractors who are being paid to write communication software and prove the software doesn't leak information outside certain components. Software modelling and generation has progressed leaps and bounds over the last ten years, and will probably continue to do so. The people behind OOA are busy pushing a model driven architecture, whereby you model software, and generate the code from that, continually tweaking the models, adding in lower level details (right down to source code hooks) and regenerating the binary. The amount of education required to keep an engineer current would be so large that even if they offered such a license, there'd be no takers!
"The problem is that the USA are full of ambulance-chasing lawyers, and if Apple did what appears the reasonable thing to do (free upgrade), then you can be guaranteed that someone would start a class action lawsuit against Apple."
Then where's the shareholder class action against Steve Jobs? Here we have a touchy case on something the public is frustrated about: executive compensation. Yet there's no shareholder outrage. You want to know why? Because investors are making good returns. Would the FTC really start a case against Apple for offering this upgrade for free? I think not. My best guess is that this is intended to be a loud complaint about the FTC and other regulations. The fact that they'll roll it out with the next upgrade is amusing, and suggests that they don't have much to lose by putting it off for a while. But really, I can't see how their logic doesn't apply to security upgrades. I hope Sarbanes or Oxley clarify some time.
The worry is that your password can be stolen and used later, not nessecarily while your computer is even on. That dongle isn't going to be easily cracked, even if you have Admin access to the computer. It's likely to use some sort of challenge response system, where you issue a challenge, and the answer contains both the correct answer and whatever additional information you'd like to forward on protected, which can also be encrypted based on the challenge. If I were to try making my own, I'm sure the NSA could hack it, but your average keylogger author might have a challenge outside his grasp. And at Blizzard's level, they could just hire RSA Corp to do it or evaluate or whatever. At which point the keylogger is almost certainly foiled, and even the NSA has a run for their money.
Seems sufficiently safe for what is, in the end, still just a video game.
Slashdot Mirror
He wasn't trying to make it look like he was downloading material. To continue your analogy, making a substance that smells and looks like pot would be sending false data to the torrent. He didn't do that. His program is observational in nature. It's a bit like calling the cops in the guy with the camera in the back alley because you think its suspicious. It doesn't transfer anything, it doesn't accept anything from seeds. Even if you tried sending garbage, you'd be caught by the md5 hashes and blocked / banned.
If your goal is to examine the nature of torrents and how different goals result in different patterns, a tool like this is vital. It's not at all straightforward that it's unethical or illegal to connect to the swarm to gather user statistics.
Sure, I got the joke. The bit about the TLAs just irked me. And you knew someone had to reply with a straight answer so I figured "Why not me?" ;)
I've heard that disney bought up florida land under a secretive context, and it was quite a bruhaha at the time when it was discovered that they were building a resort there. Perhaps google is worried that if word gets out they're making a large investment there, that someone will start looking for where google would want to locate (perhaps a large unused building located in a large tech heavy town, say an old AOL building?)
Page rank was/is hands above its peers at introduction, but it's not pagerank that makes google spectacular. It's about making pagerank fast. Google successfully improved web search, and then made it scalable both in number of users and in pages indexed.
Open source is good at providing new ideas, but high performance doesn't have a huge market for users. Many of Google's most important optimizations require a special configuration and hardware, like thousands of computers laden with RAM, power etc. As it happens, wikipedia often times cannot handle it's users demands. Their search feature is sometimes handed off to google. Lets also not forget the amount of data required to be stored. Sure, you can download tons of websites, but you can't discard everything; you'll need the url, relevant links, and context, excerpts to show users, etc.
Of course, all this pessmism hasn't stopped people from trying. If you discover a list of sites running Nutch, you'll notice most are niche sites, either offering to search a single website, or subtopic. The only general web search deployment I found was fast but crappy.
This comment hurts my brain. You bitch about TLAs but they're not even acronyms. Do you even know what TLA stands for?!
Lockless means it's doing something without locking everyone else from the data. Sometimes this means optimistic resolution (everyone try, and if it looks like it screwed up, try again!), sometimes it means keeping local read copies, sometimes it means something new and/or crazy. Lockless approaches are used when you have data that lots of threads must share but efficiency concerns or non-blocking requirements force you away from simply using a lock and blocking when someone else is playing with the data.
A radix-tree, as opposed to "radix-free", is a data structure used in certain applications, with operations dependent on the length of the key rather than the amount of data stored. In 2.6.20 (and others), it's used to organize some information about the page cache.
This code is associated with the RCU, which you may recall is part of an SCO lawsuit. If you're interested in any other feature or changes, the kernel newbies site is instrumental!
I haven't shopped at ebay in a long time, but apparently one can withdraw a bid. So it should come as no surprise that scammers use this feature to eliminate the risk of overbidding. Meanwhile, the consumer is competing with demand that doesn't actually exist. Apparently it's popular enough that ebay offers a Second Chance feature, that lets a seller offer the second highest bidder the option to buy at their highest losing bid should the winner fail to complete the transaction.
I guess as long as ebay is interested in more sales and higher prices, there's too much incentive to address the problem poorly.
I can write an O(1) scheduler too. FIFO comes to mind...
"Oh, the horrors! I've had this happen to me many times, as well, especially on Linux. Not even a kill -9 would get rid of the wedged process. Why is there even such a thing as "non-interruptible sleep"? If I don't need the process anymore, I should be able to get rid of it no matter what."
I haven't yet read the internals of the linux kernel, but I understand that disk dma requests are a major source of non-interruptible sleep. I imagine it has to do with locking the RAM the disk is going to write to safe from other processes. But from my studies of MINIX, I recall that the DMA controllers were a bit limited, so its possible DMA in linux is done to some buffer and then copied into userspace. This would make my whole argument invalid =(
Don't worry. It can't be lock bumped: the same key opens every machine.
Fantasy land material. Wal-mart's data centers would eat MS products alive. Recall that every transaction is being logged there. About seven years ago, my university recieved a donation of one of their district processing mainframes: something like an 82 way pentium 2 setup. Fantastic sounding stuff, but it was a) too slow for their (regional) needs, and b) too damned hard to make fast (NUMA).
If Walmart was dissatisfied with Linux, somehow I think Windows would be their last pick. Which makes me wonder, what are they using now? Linux? Solaris?
Usually the time of day metering is for industrial users, who buy ridiculus amounts ( think robotic spot welders)
As far as I know, the GMA 9xx series is a couple generations behind, performance wise. It should play quake3 and UT2k4 just fine, but it seems to have trouble with the Doom 3 engine, and I suspect the new UT engine will also be unplayable. On the windows side, it doesn't work with halflife2 either. Seems the most likely kind of game to fail is a new FPS. But I hear aero and Xgl/AIGLX work fine, so you may be satisfied with the current Intel offerings. The wikipedia page seems like a good place to start researching if you're still interested.
Why does liscencing a patent require closed source drivers? If I recall correctly, the patent system requires disclosure as reciprocation for limited monopolies. I believe there are open source 3d drivers for the savage3 chipset, which is strange given the circumstances claimed.
Sadly, while your energy production peaks coincide with peak demand, you're still paid a flat rate for the production. I hear that power companies charge some high energy users more during peak hours, and less during the night / troughs of low demand. Paying out more for energy during peak hours would probably improve the incentives to go solar, but I guess the cost of monitoring this would be huge compared to reading a number every month.
I don't know what it's doing on a site called "unrealnation", but there is a mirror, I think.
The news here isn't that they're using x86 architecture. It's that they're using Intel chips. They already have AMD Opterons for sale, and they're adding Xeons. And Solaris 10 does run on x86. Some see it as a concession that their server CPU designs are little more than a niche market. That diagnosis is probably correct, and if Sun wants to ever dig themselves out of the "Sun is dying" meme, they'll have to start taking advantage of the fact that their competitors are engaged in a price war, one that's also cutting into profits. Sun can still pride themselves on quality server hardware, support contracts and integration, even if they aren't the designer behind most their chips.
Meanwhile, their Niagra still has some niche applications, and will grow as software is designed for dual and quad core chips. If Niagra does what they say it will, people will be forced to consider one Niagra unit versus 6 Xeon servers. Xeons may have fallen in price recently, but they're still not cheap, so that's a calculus that Sun might win some day.
They're getting into the business because they have tons of money to invest, they already know the system well enough, and the only significant competition is PayPal. They've been offering "pay services" since about 2000-ish. Remember Adwords? The whole thing's automated and doesn't rely on paypal or other services that compete with google checkout. Way older than charging for Picasa / Google Earth.
They offered the money for one simple reason: adoption rates. I can't even begin to describe the amount of money paypal threw at the altar of adoption rates. By giving money away, they're quickly getting merchants who need an edge over their competitors, or are just worried about their competitors doing the same. And customers have the usual incentives to take up the deal. The reason adoption rates matter is that the value of their service is based on the number of possible transactions that an individual can participate in. The more merchants using it, the more likely you'll want to sign up for a google checkout account, even without the discount offers. And the more consumers with accounts, the more sense it makes to put forth the effort to integrate google checkout into your website. It's a virtuous cycle, but Google needed to give it a push start. I'm not sure one needs to resort to a data mining conspiracy to justify their actions here.
"Linus may fix an issue with the VM subsystem, but usually won't be the person to fix a problem with ext2"
It's kinda interesting that you bring this up, becuase he recently went through and fixed a data loss problem caused by interaction between VM and ext2. Just because someone else is the maintainer of it doesn't mean he can't go through and debug it like anyone else. But I think that there are plenty of people who would like to go through the nvidia source code. They may not all be experts, but I have a sneaking suspcion that the nvidia team isn't either. At first, you'll probably see a few code critiques, during which the module maintainers may be exposed to "kernel best practices" rants. But for all their internal knowledge, there's likely enough documentation in the drivers alone to more than double the efforts at improving nvidia drivers after 6 to 12 months. For example, the nouveau project is already making progress towards an open source driver for nvidia 3d chipsets, without anything more than the binary drivers, cards, and helpful users. Even if a tragically small percentage of OSS developers are able to help nouveau, it's still more than nvidia has dedicated to it.
"Judge Wells asked IBM to help SCO out in any way he could."
Maybe this is why I'll never go to law school, but I'd probably respond to that at that point by placing a loaded gun on the table and tell them "Kill yourself. Now."
On the contrary, when taking the article at face value, I think that the whole intent is to make it *easy* for Microsoft to implement the first time, because it's already done. This is of course backwards from how a community standard should work, it should be an effort that is repeatable. Instead we have whatever crap their contractors turned in, with apparent flaws turned into requirements. I doubt even Microsoft could write a second compatible handler for this document format. Well, perhaps instead of "on the contrary" I should say "to clarify"; the standard appears to be designed so that any implementation but the original is near impossible.
It sounds like pretty much like business as usual for MSFT, although describing in 6,000 pages how hard it would be to create an interoperating product is new. Their format is the standard, even the flaws that they didn't fix before release.
Your tears are so yummy and sweet!...Oh, the tears of unfathomable sadness!
-Cartman
The engineering profession works mostly the same way in the US. The Texas deal was them trying to jumpstart a Software Engineering field. If you look at the specializations, there's a huge preference for building codes. Fire safety engineering. Civil Engineering. There's no computer chip engineering specialty. Basically, the definition of "engineer" to them is someone who takes thermodynamics.
But if you recall that engineering comes from designing new and useful things, rather than specifically building a bridge or an engine, the case for Software Professional Engineers gets strong. There are tons of public systems in use today that rely on software, but the state currently has no means of assessing quality. The reason the US hasn't had to re-examine this in great detail is the exception clause: if you're work is for a specific employer "in industry" then the company takes liability, not you. Of course, most software comes with "NO GUARENTEE FOR ANY PARTICULAR USE" etc. Somehow, the DoD manages to survive simply by assessing bidders, although I hear the Grand Challenge was partly started as an alternative to the failures DARPA had in getting products of value from industry.
It sounds like there is a valid case to be had for professional licensing of software developers. Whether that term is "engineer" is moot; the end result would likely look and operate in the exact same way. The strongest counter-argument, and the one I never hear from people invested in the Professional Engineer licensing program, is that software is too young to encode a lasting best practices and design techniques. Five years ago, you wouldn't likely have heard of unit testing. Verification and validation of software was something PhDs did; now I know of a few defense contractors who are being paid to write communication software and prove the software doesn't leak information outside certain components. Software modelling and generation has progressed leaps and bounds over the last ten years, and will probably continue to do so. The people behind OOA are busy pushing a model driven architecture, whereby you model software, and generate the code from that, continually tweaking the models, adding in lower level details (right down to source code hooks) and regenerating the binary. The amount of education required to keep an engineer current would be so large that even if they offered such a license, there'd be no takers!
"The problem is that the USA are full of ambulance-chasing lawyers, and if Apple did what appears the reasonable thing to do (free upgrade), then you can be guaranteed that someone would start a class action lawsuit against Apple."
Then where's the shareholder class action against Steve Jobs? Here we have a touchy case on something the public is frustrated about: executive compensation. Yet there's no shareholder outrage. You want to know why? Because investors are making good returns. Would the FTC really start a case against Apple for offering this upgrade for free? I think not. My best guess is that this is intended to be a loud complaint about the FTC and other regulations. The fact that they'll roll it out with the next upgrade is amusing, and suggests that they don't have much to lose by putting it off for a while. But really, I can't see how their logic doesn't apply to security upgrades. I hope Sarbanes or Oxley clarify some time.
"I think they realize that very few people are probably interested in 802.11n, because few have equipment to work with it."
Apparently in some cases, the equipment is already there, waiting to be upgraded, perhaps without the customer even realizing it.
The worry is that your password can be stolen and used later, not nessecarily while your computer is even on. That dongle isn't going to be easily cracked, even if you have Admin access to the computer. It's likely to use some sort of challenge response system, where you issue a challenge, and the answer contains both the correct answer and whatever additional information you'd like to forward on protected, which can also be encrypted based on the challenge. If I were to try making my own, I'm sure the NSA could hack it, but your average keylogger author might have a challenge outside his grasp. And at Blizzard's level, they could just hire RSA Corp to do it or evaluate or whatever. At which point the keylogger is almost certainly foiled, and even the NSA has a run for their money.
Seems sufficiently safe for what is, in the end, still just a video game.