I'm not sure exactly what software they used for this benchmark, but the G5's AltiVec vector unit is several orders of magnitude faster than the SSE2 vector unit on the Xeon. If they used software that is optimized for AltiVec, it doesn't surprise me in the slightest that the G5 beat out the Xeon hands down. What would surprise me is if this difference was in unoptimized code -- ie, code that doesn't use the vector units.
I suppose I should have linked to the source where I copied that quote from, since it wasn't directly linked from the front page -- here. The issue isn't that it is in beta, or is a research project, but that the project has projected the illusion of being an anonymous and secure way to share controversial, or even possibly illicit, material. There are many people who have put a lot on the line on the basis that freenet meets these standards. As in, they could go to jail simply because of the information they have put out on freenet.
Freenet is free software designed to provide a forum where information can be published and consumed without fear of censorship. It does this by providing a completely decentralized, and robust way that people can publish and read information anonymously. Freenet grew out of a paper I wrote while still a student at Edinburgh University.
Sounds like the canary has changed its tune, eh? Now freenet is a research project, not a 'forum where information can be published and consumed without fear of censorship.' Although I always respect a developer that wants to go back and fix bugs with a system before moving to another release (or I suppose in this case, after moving to another release), the email from Ian Clarke sounds downright aggorant -- you can address points about bugs without telling someone to go use another network. I don't use freenet, so it doesn't really affect me, but I definately feel sorry for those who do/did.
Although I agree with the idea of your post, there are several dorms at the University of Florida with full OC3 connections to them -- the students aren't exactly bandwidth starved:P. I'm not sure what kind of upstream billing arangement UF has, but most likely the reason they want to get rid of P2P is because it uses a tremendous amount of upload bandwidth, which is usually quite a bit more expensive than download bandwidth.
Its fairly wierd actually. Microsoft also has a program where if a group of college students want to work on some computer science/engineering related project but do not have the necessary resources, they will help provide funding for it -- even if it is an open source or GPL'ed project. I suppose Microsoft's help with Xen is an extension of this program. I don't think their interests are totally philanthropic either -- if Xen happens to hit it off and become widespread, they're definately going to want it to run Windows.
Indeed, but that is what we must deal with to have such wonderful innovations as Clippy(tm) for the toaster.
"It looks like you're making toast! Would you like help?"
The problem is, you assume there is some arbitrary majority that agrees on all issues. If there was a 51% portion of the population that agreed on everything, or even on most things, then what you suggest would indeed happen. However, people have a wide range of opinions on a wide range of different issues, so although you may be able to unite a majority behind one issue, you may alienate them with another. This is why democracy continues to work even in a society the size of the United States. Why do you think we still get both Democrats and Republicans in office -- if there really was some majority that voted the same for everything, then they would obviously choose one political party and stick with it. Given that a goodly number of Americans have participated in filesharing (I haven't found any statistics, but I would feel safe saying that at least a majority have been exposed to it, even if they haven't used it themselves) if the majority was the ruling factor on all issues then there probably would/will be legislation concerning filesharing. This is a minor issue from the perspective of the government (not to mention most of them don't read Slashdot -- if they did, they would most likely presume that SCO, the RIAA, and Microsoft were the top three issues in the country;) ), and I would be appalled if they didn't treat it as such. The government will get around to making filesharing laws, just like it has begun to get around to making spam laws. Its all about priorities.
You want to know why democracy doesn't always work in this country? <rant> Its because most people don't get off their lazy ass once every couple of years, learn a little bit about whats going on in politics, and then vote. Barely more than 50% of the people in this country vote in presidential elections, and far fewer vote for state and local elections. For those people who don't vote, you reap what you sow -- you gave up your right to have a say in this country when you didn't vote.</rant>
I agree with you, but one would think that MIT students wouldn't be the type of people that would get a computer virus. I mean, really, these are some of the best and brightest engineering students in the world, and they didn't patch their systems for blaster?
I wonder how they actually ensure that you reformat your computer, however (which just opens it up to any security holes the user may actually HAVE patched)... do they force you to reformat in front of network personell? I can't imagine they have the kind of staff to sit around there and wait while each infected student reformats and reinstalls their computer.
The Quake/Quack scandal involving ATi never existed, of course.
I'm not sure what engine Half-Life 2 is based off of, but given that Half-Life was based off of the Quake engine it wouldn't surprise me if Half-Life 2 was based of the Quake 3 engine. Perhaps this scandal is really just an extention of the Quake optimization issue.
Thank you. If more sysadmins actually monitored their servers for malicious activity, we wouldn't have the thousands upon thousands of DDoS trojans out there on the Internet today. Too many sysadmins think you can just install all of the software, get it working right, and then never look at their server(s) again.
The US is actually trying to reduce the number of nuclear power plants, which is why you see few advances in nuclear power in the United States. Most state governments have been unable to sell the idea of "nuclear power in your backyard" to their citizens. Add to this the slew of new security regulations placed on nuclear power plants since 9/11/2001 and no one is willing to build any new plants or implement any new technology. Without nuclear power, though, I don't see how the US plans to power itself over the next 20-50 years when major fossil fuel supplies start to empty out.
It depends. Perhaps you left a door unlocked, and because he told you that it was unlocked you were able to fix it. Then when the real thieves came next week you would be protected against them.
The assertation that there are only criminals and law-abiding citizens is a horrible bifurcation. If you speed when you're taking a loved one to the hospital, are you a criminal? If you kill someone in self defense or in defense of another, are you a criminal? Maybe this guy's actions aren't as noble as these examples, but he did no harm and from the words of the system admins he did a lot of good.
Unfortunately, you failed to come up with such a creative name as "Billy Goat" for your project. Who can resist software called "Billy Goat"? Perhaps you can call your project "She-Buffalo" and you'd have a chance!
There are many. One of the advantages of Qt is that it provides a common interface to X-Windows, MS Windows, and Mac OS GUI programming. Qt is entirely independent of KDE, the only reason an application would be bound to KDE is if it utilizes the KDE extensions to Qt.
I wouldn't exactly use Sendmail as an example of a secure mail transfer system. Perhaps nowadays it isn't bad, but back in the day it was famous for the number of worms that spread through it:P.
There are ways for this virus to get through a firewall. For example, the virus entered our corporate network through a user's VPN connection to an infected network. Luckily very few machines on the network had not been patched, so the impact was minimal.
Incorrect. Poorly designed software is getting way to complex for human management in developing bug-free code. Code that uses encapsulated, individually testable objects and the full features of today's object-oriented languages is easily managed. Objects in a debug build should be able to verify their own integrity, raising alerts when things aren't as they should be.
Think about it. A large building has probably far more millions of components than most code has lines I don't see Civil Engineers going around saying "yeah, our buildings pontaneously collapse every so often, but that's part of making buildings!" They don't even get to test their construction before it is released, as it is easily possible to do with software. However, the design process of making a building is far more rigorous because it has to be. People just get lazy when writing software because there aren't any dire consequences when it doesn't work right.
Bugs are expected as part of the development and testing process -- nobody's perfect. However, release software should be virtually bug-free. By no means should system crashes be *routine* and *expected* as it is in the system described by this article. The IT department in the company I work for does not have a code maintenance division -- we have never needed one. Although our largest project has probably been only around 100,000 lines of code (as opposed to two million) the beauty of object oriented design is that it scales to the project -- if and only if implemented properly!
You have to buy 2.5% more stuff, but since that stuff is cheaper, you are more or less spending the same amount of money -- essentially, you are getting more bang for your buck. It is definately possible, although highly unlikely, for a capitalist economy to reach a state of equilibrium. People would get more "stuff" each year, yet have a constant income and a constant employment rate. In the real world, this is unfeasable due to the unpredictability of the masses, but it is nevertheless possible.
This is sad for Microsoft. Prejudices aside, I simply cannot understand how open source software can compete with Microsoft's software on a large scale. MS has the resources to hire the best, brightest, and most innovative minds in the industry, but it is losing ground to software written as a hobby. Its like some guy making movies in his backyard for fun -- and competing with Hollywood. Sounds like Microsoft has a seriously flawed R&D department to me...
Yes, but a state transition requries an increase or decrease in the atom's energy level. Electricity doesn't necessarily increase or decrease the energy level of an atom -- if you apply electricity to a random object, it doesn't just start emitting light. The significance of this discovery is that the atom will emit light when a regular electric current is applied to it.
Agreed, especially on the second point. Biohistory has shown that species that are able to mutate quickly and have wide genetic diversity are the ones that will survive in the longterm. Take bacteria for example. Phenomenally fast mutation rate and very genetically diverse. Thats why there are trillions of them, and they are able to survive in pretty much any area of the earth, including areas with extremely high temperatures or no oxygen. With a highly specialized genetic sequence, a disease such as smallpox would decimate human population. Instead of spreading rapidly, killing 20% of the people it infects and becoming endemic, it would remain pandemic until population was too sparse to support its spread. Pretty much everyone understands the theory of natural selection -- woe be unto humanity should natural selection choose a genetic sequence that we engineered out of existance.
Have you ever gotten a free email address at Yahoo? I've had an email address there for a little over 4 months and have received a grand total of 2 spam emails. This doesn't sound spam-friendly to me, as an equivalent account at a provider such as hotmail will land you 10 spam emails within the first 24 hours. As a service provider, they have absolutely no interest in being spam-friendly -- it only sucks up their bandwidth and their time.
As for the criminal activity, I haven't read anything about it but it is always possible. However, it is against a service provider's best interests to actively search for and shut down any possible criminal acts. If a company proves that they are able at any point in time to filter or find criminal-type acts it becomes a serious liability should they ever be taken to court for hosting illegal content. Given that Yahoo/Geocities hosts thousands of users' pages, it isn't realistic for them to monitor each one individually to make sure every individual site is always following their terms of service. Instead, they remove pages if users notify them of the violations (like the DMCA's provision for removing copyrighted material).
I have an HP LaserJet that is 5 years old and still going strong. For color printers, I've had a Lexmark, a Cannon, and two HP printers (one of which I'm currently using, the others have worn out). Supposing I had bought all HP printers, they would have made a lot more money of off all of those color printers I bought than that old laser printer which still works wonderfully today.
I've noticed the same thing with computers. I have an old IBM Aptiva -- about 7 years old -- that, although slow, is still going strong today. I haven't even had to replace the CMOS battery in it! I had one Compaq that burned its motherboard out within two years, and a Gateway laptop that shuts off if you so much as tap it while its running a CPU-intensive program (luckily, although Gateway computers aren't really good, their tech support is great and they're happy to take it in and replace whatever is necessary).
The real problem is that computer geeks don't make up the majority of computer and peripheral consumers. There are far more 'Average Joes' and businesses out there than there are computer geeks. 'Average Joes' don't know enough about technology to make an educated computer or printer choice -- they buy what's cheapest or what (and I quote) 'has the most Gigahertz.' Same thing with businesses -- in an effort to cut costs, purchasing managers may choose to buy a cheaper printer or computer and let someone else deal with the maintenance issues. It looks good on them to cut costs, and most likely no one will think to blame them when the printer or computer breaks down a year later.
Not exactly. You put information out on the web for the purpose of it being accessible -- if you didn't want it to be accessible, you wouldn't put it on the web. You don't set up an email address for the purpose of getting spam (hopefully not, at least). And if you consider penis enlargement and other such spam 'informative and useful mail', well, no comment:P.
All of these people complain about slashdot linking to pages, but you won't see them stop clicking on the links to help solve the problem. Although I do believe they could be a little more considerate about linking to small personal pages, it is the responsibility of the web server's administrator to set up policies to avoid bandwidth overusage. There are hundreds of ways a website can get its bandwidth eaten up like crazy, including being in the top 10 on a google search, being linked by any large news site, or just rampantly (un)lucky word-of-mouth. If the system administrator of a web server has done nothing to compensate for cases of usage spikes, it is not slashdot's place to do it for them.
Slashdot Mirror
I'm not sure exactly what software they used for this benchmark, but the G5's AltiVec vector unit is several orders of magnitude faster than the SSE2 vector unit on the Xeon. If they used software that is optimized for AltiVec, it doesn't surprise me in the slightest that the G5 beat out the Xeon hands down. What would surprise me is if this difference was in unoptimized code -- ie, code that doesn't use the vector units.
I suppose I should have linked to the source where I copied that quote from, since it wasn't directly linked from the front page -- here. The issue isn't that it is in beta, or is a research project, but that the project has projected the illusion of being an anonymous and secure way to share controversial, or even possibly illicit, material. There are many people who have put a lot on the line on the basis that freenet meets these standards. As in, they could go to jail simply because of the information they have put out on freenet.
# What is Freenet?
Freenet is free software designed to provide a forum where information can be published and consumed without fear of censorship. It does this by providing a completely decentralized, and robust way that people can publish and read information anonymously. Freenet grew out of a paper I wrote while still a student at Edinburgh University.
Sounds like the canary has changed its tune, eh? Now freenet is a research project, not a 'forum where information can be published and consumed without fear of censorship.' Although I always respect a developer that wants to go back and fix bugs with a system before moving to another release (or I suppose in this case, after moving to another release), the email from Ian Clarke sounds downright aggorant -- you can address points about bugs without telling someone to go use another network. I don't use freenet, so it doesn't really affect me, but I definately feel sorry for those who do/did.
Although I agree with the idea of your post, there are several dorms at the University of Florida with full OC3 connections to them -- the students aren't exactly bandwidth starved :P. I'm not sure what kind of upstream billing arangement UF has, but most likely the reason they want to get rid of P2P is because it uses a tremendous amount of upload bandwidth, which is usually quite a bit more expensive than download bandwidth.
Its fairly wierd actually. Microsoft also has a program where if a group of college students want to work on some computer science/engineering related project but do not have the necessary resources, they will help provide funding for it -- even if it is an open source or GPL'ed project. I suppose Microsoft's help with Xen is an extension of this program. I don't think their interests are totally philanthropic either -- if Xen happens to hit it off and become widespread, they're definately going to want it to run Windows.
Indeed, but that is what we must deal with to have such wonderful innovations as Clippy(tm) for the toaster. "It looks like you're making toast! Would you like help?"
Eh? Ctrl-alt-backspace kills X, unless you have DontZap in your X config file.
You want to know why democracy doesn't always work in this country? <rant> Its because most people don't get off their lazy ass once every couple of years, learn a little bit about whats going on in politics, and then vote. Barely more than 50% of the people in this country vote in presidential elections, and far fewer vote for state and local elections. For those people who don't vote, you reap what you sow -- you gave up your right to have a say in this country when you didn't vote.</rant>
I agree with you, but one would think that MIT students wouldn't be the type of people that would get a computer virus. I mean, really, these are some of the best and brightest engineering students in the world, and they didn't patch their systems for blaster? I wonder how they actually ensure that you reformat your computer, however (which just opens it up to any security holes the user may actually HAVE patched)... do they force you to reformat in front of network personell? I can't imagine they have the kind of staff to sit around there and wait while each infected student reformats and reinstalls their computer.
I'm not sure what engine Half-Life 2 is based off of, but given that Half-Life was based off of the Quake engine it wouldn't surprise me if Half-Life 2 was based of the Quake 3 engine. Perhaps this scandal is really just an extention of the Quake optimization issue.
Thank you. If more sysadmins actually monitored their servers for malicious activity, we wouldn't have the thousands upon thousands of DDoS trojans out there on the Internet today. Too many sysadmins think you can just install all of the software, get it working right, and then never look at their server(s) again.
The US is actually trying to reduce the number of nuclear power plants, which is why you see few advances in nuclear power in the United States. Most state governments have been unable to sell the idea of "nuclear power in your backyard" to their citizens. Add to this the slew of new security regulations placed on nuclear power plants since 9/11/2001 and no one is willing to build any new plants or implement any new technology. Without nuclear power, though, I don't see how the US plans to power itself over the next 20-50 years when major fossil fuel supplies start to empty out.
The assertation that there are only criminals and law-abiding citizens is a horrible bifurcation. If you speed when you're taking a loved one to the hospital, are you a criminal? If you kill someone in self defense or in defense of another, are you a criminal? Maybe this guy's actions aren't as noble as these examples, but he did no harm and from the words of the system admins he did a lot of good.
Unfortunately, you failed to come up with such a creative name as "Billy Goat" for your project. Who can resist software called "Billy Goat"? Perhaps you can call your project "She-Buffalo" and you'd have a chance!
There are many. One of the advantages of Qt is that it provides a common interface to X-Windows, MS Windows, and Mac OS GUI programming. Qt is entirely independent of KDE, the only reason an application would be bound to KDE is if it utilizes the KDE extensions to Qt.
I wouldn't exactly use Sendmail as an example of a secure mail transfer system. Perhaps nowadays it isn't bad, but back in the day it was famous for the number of worms that spread through it :P.
There are ways for this virus to get through a firewall. For example, the virus entered our corporate network through a user's VPN connection to an infected network. Luckily very few machines on the network had not been patched, so the impact was minimal.
Incorrect. Poorly designed software is getting way to complex for human management in developing bug-free code. Code that uses encapsulated, individually testable objects and the full features of today's object-oriented languages is easily managed. Objects in a debug build should be able to verify their own integrity, raising alerts when things aren't as they should be. Think about it. A large building has probably far more millions of components than most code has lines I don't see Civil Engineers going around saying "yeah, our buildings pontaneously collapse every so often, but that's part of making buildings!" They don't even get to test their construction before it is released, as it is easily possible to do with software. However, the design process of making a building is far more rigorous because it has to be. People just get lazy when writing software because there aren't any dire consequences when it doesn't work right. Bugs are expected as part of the development and testing process -- nobody's perfect. However, release software should be virtually bug-free. By no means should system crashes be *routine* and *expected* as it is in the system described by this article. The IT department in the company I work for does not have a code maintenance division -- we have never needed one. Although our largest project has probably been only around 100,000 lines of code (as opposed to two million) the beauty of object oriented design is that it scales to the project -- if and only if implemented properly!
You have to buy 2.5% more stuff, but since that stuff is cheaper, you are more or less spending the same amount of money -- essentially, you are getting more bang for your buck. It is definately possible, although highly unlikely, for a capitalist economy to reach a state of equilibrium. People would get more "stuff" each year, yet have a constant income and a constant employment rate. In the real world, this is unfeasable due to the unpredictability of the masses, but it is nevertheless possible.
This is sad for Microsoft. Prejudices aside, I simply cannot understand how open source software can compete with Microsoft's software on a large scale. MS has the resources to hire the best, brightest, and most innovative minds in the industry, but it is losing ground to software written as a hobby. Its like some guy making movies in his backyard for fun -- and competing with Hollywood. Sounds like Microsoft has a seriously flawed R&D department to me...
Yes, but a state transition requries an increase or decrease in the atom's energy level. Electricity doesn't necessarily increase or decrease the energy level of an atom -- if you apply electricity to a random object, it doesn't just start emitting light. The significance of this discovery is that the atom will emit light when a regular electric current is applied to it.
Agreed, especially on the second point. Biohistory has shown that species that are able to mutate quickly and have wide genetic diversity are the ones that will survive in the longterm. Take bacteria for example. Phenomenally fast mutation rate and very genetically diverse. Thats why there are trillions of them, and they are able to survive in pretty much any area of the earth, including areas with extremely high temperatures or no oxygen. With a highly specialized genetic sequence, a disease such as smallpox would decimate human population. Instead of spreading rapidly, killing 20% of the people it infects and becoming endemic, it would remain pandemic until population was too sparse to support its spread. Pretty much everyone understands the theory of natural selection -- woe be unto humanity should natural selection choose a genetic sequence that we engineered out of existance.
Have you ever gotten a free email address at Yahoo? I've had an email address there for a little over 4 months and have received a grand total of 2 spam emails. This doesn't sound spam-friendly to me, as an equivalent account at a provider such as hotmail will land you 10 spam emails within the first 24 hours. As a service provider, they have absolutely no interest in being spam-friendly -- it only sucks up their bandwidth and their time. As for the criminal activity, I haven't read anything about it but it is always possible. However, it is against a service provider's best interests to actively search for and shut down any possible criminal acts. If a company proves that they are able at any point in time to filter or find criminal-type acts it becomes a serious liability should they ever be taken to court for hosting illegal content. Given that Yahoo/Geocities hosts thousands of users' pages, it isn't realistic for them to monitor each one individually to make sure every individual site is always following their terms of service. Instead, they remove pages if users notify them of the violations (like the DMCA's provision for removing copyrighted material).
I have an HP LaserJet that is 5 years old and still going strong. For color printers, I've had a Lexmark, a Cannon, and two HP printers (one of which I'm currently using, the others have worn out). Supposing I had bought all HP printers, they would have made a lot more money of off all of those color printers I bought than that old laser printer which still works wonderfully today. I've noticed the same thing with computers. I have an old IBM Aptiva -- about 7 years old -- that, although slow, is still going strong today. I haven't even had to replace the CMOS battery in it! I had one Compaq that burned its motherboard out within two years, and a Gateway laptop that shuts off if you so much as tap it while its running a CPU-intensive program (luckily, although Gateway computers aren't really good, their tech support is great and they're happy to take it in and replace whatever is necessary). The real problem is that computer geeks don't make up the majority of computer and peripheral consumers. There are far more 'Average Joes' and businesses out there than there are computer geeks. 'Average Joes' don't know enough about technology to make an educated computer or printer choice -- they buy what's cheapest or what (and I quote) 'has the most Gigahertz.' Same thing with businesses -- in an effort to cut costs, purchasing managers may choose to buy a cheaper printer or computer and let someone else deal with the maintenance issues. It looks good on them to cut costs, and most likely no one will think to blame them when the printer or computer breaks down a year later.
Not exactly. You put information out on the web for the purpose of it being accessible -- if you didn't want it to be accessible, you wouldn't put it on the web. You don't set up an email address for the purpose of getting spam (hopefully not, at least). And if you consider penis enlargement and other such spam 'informative and useful mail', well, no comment :P.
All of these people complain about slashdot linking to pages, but you won't see them stop clicking on the links to help solve the problem. Although I do believe they could be a little more considerate about linking to small personal pages, it is the responsibility of the web server's administrator to set up policies to avoid bandwidth overusage. There are hundreds of ways a website can get its bandwidth eaten up like crazy, including being in the top 10 on a google search, being linked by any large news site, or just rampantly (un)lucky word-of-mouth. If the system administrator of a web server has done nothing to compensate for cases of usage spikes, it is not slashdot's place to do it for them.