I feel obliged to point out that in addition to being wrong about the motivations, he's also wrong that Tanenbaum thought that monolithic kernels were slow. Everyone knew that monolithic kernels were faster - fewer context switches, no message-passing overhead. His claim was (and is) that microkernels were better, not that they were faster. And the success of Linux really doesn't address the fundamental design and architetural differences between micro and macrokernels at all.
It's a contract in an abstract meaning of the word, like you might say "social contract", but in (US) legal terminology, the difference is important. And since the grandparent was asking about breach of contract damages, thats the usage I was addressing.
Nonsense. Sure, the Earth (obviously) hasn't been colonized because no one has tried, but claiming thats proof that extraterrestial life doesn't exist is like claiming that because noone has settled a given spot on the ocean floor theres no evidence of life on Earth. At the very best, you can presume that interstellar travel is difficult and/or dangerous (certainly supportable) and that intelligences zipping about like Star Trek don't exist.
You're defining successful intelligent life as "life which would have colonized Earth" and noting that because it didn't happen, there is no successful life. A species that managed to colonize *any* other planets is going to be one much more advanced than ours, and a species that actually maintained some sort of interstellar society far, far so. If such a species existed, they would neccesarily need forms of communication and travel we aren't aware of and cannot detect, and yet they still could easily be sufficently distant from us to not have colonized Earth. Note that I'm not suggesting such a species exists, and that while they *could*, it's reasonable to work from the assumption that they don't. But going from "they probably don't" to "they must not" exist is silly. Think of it as religion if you want - at this point, it's a matter of faith or assumption, not of evidence. Certainly, if other intelligences do exist, they are irrelevent to us this point (barring sci-fi ideas that they know about Earth and are intentionally avoiding it for whatever reason).
If you want to learn how to *program*, then you need an itch you want to scratch. Taking a programming course, or reading a book and working the samples will teach you about programming, and maybe computer science (depending on the book), but won't teach you how to program. And learning about programming, or computer science, or computer hardware and engineering might be what you want - it's an interesting topic. But if what is bothering you is something like "this program sucks, I need something better" or "it'd be nice this program did this thing", or "I need a way to automate this because it's annoying", then you want to get a book on a appropriate language (I heavily recommend against Perl or C++, try Python or Ruby for local stuff) and start hacking on it until your problem goes away.
This is totally retarded, and the OPs "problem" is equally retarded. If you have a problem with "other people" modifying "your code", then you need to fix yourself, not your database.
It's copyright violation, not breach of contract (the GPL is not a contract). An amusing point, that I don't believe has ever been explored, is that many GPL violations are large scale, *commercial* violations - that is, they expose the violator to criminal prosecution as well as civil liability.
I'd have to disagree. Maybe all the NICs *you* run across, but few of the onboard ones that are so common these days, and none of the Gigabit ones that are more and more popular do. Of course, a NIC that works in a PC you're likely to want to run DOS on will likely have a DOS driver somewhere.
See http://tiny.seul.org/en/ for someone who's already done the work of stripping down Linux to run on an 8meg 486.
Presuming that by "russian piracy sites" you mean sites that offer downloads legal in Russia, then there is nothing illegal about that. It's like claiming that it should be illegal for a company to pay it's workers in China less than the US minimum wage, or for a US company with a factory in Indonesia to not maintain US environmental and work hazard standards.
You're creating a false relationship here between people who think that copyright infringment isn't theft, and people who think that makes copyright infringment okay.
It's not theft, but it's not okay (there are lot of things that aren't theft that are still both illlegal and wrong), and they should comply with the GPL, and people shouldn't download stuff illegally from P2P applications.
Note that none of that means that the music industry isn't stupid for understanding the competition they face from illegal downloads and addressing it reasonably (you can be within your legal rights but still be stupid), or that it's not legitimate to legally buy your music from overseas where it's cheaper.
This "argument" relies upon an *enormous* amount of assumptions, to the point where I question the intellecual honesty of making it. How about we just say that we know of no other intelligent life in the Universe, and have no way of knowing that they exist or not at this point, and leave it at that.
However the idea that MS would use the OSI logo without approval is silly.
It's not just MS. And in any case, the whole point of trademark law and associated things is that it's not a winning proposition to rely on people to act correctly. If there were no legal consequences from using the OSI mark without permission, would you still dismiss the possibility offhand?
He doesn't seem to be saying that here, and the official OSI position is quite clear (search the OSI site for "trademark").
Whether you approve or not, OSI certainly is *a* authority on what "open source" means, and the OSI certified logo is respected for that reason. Maybe not by you, but you don't have to respect that if you don't want to. As for why someone might want to use the logo without approval, you might as well ask why someone would want to put the UL logo on thier product.
IBM has contribued a signifigant amount of code to the Linux kernel, thats all under the GPL. They've contributed to several other projects, that code is under the license of the project its donated to. They've open sourced several projects whole sale, although I'm not sure how many if any of those are under the GPL (but they are under OSI approved OSS licenses). IBM is one of the major contributors to Eclipse.
With all due respect, you don't know what you're talking about. Clustering is not some cutting edge new technology, it's well known and proven and competent administrators are perfectly capable of implementing very high reliability infrastructure using off the shelf parts. Custom kernel tweaking is something you don't need at all. In fact, it requires more specialized talent and more expensive people to run your mainframe.
The political aspects of needing to outsource operations for CYA are totally irrelevent to the technology involved - there are plenty of companies more than willing to build server farms for you. In fact, there are more of them than there are companies willing to manage your 100 million dollar mainframe for you.
Lastly, the amount of traffic implied by 3000 containers a day, while certainly enterprise class, doesn't come any where near the high end of other existing systems.
Many, many crucial business and government systems run on server clusters. It's not any sort of career suicide to do it.
Java excels where you need to interface with legacy and/or disparate systems.
This is exactly contrary to my experience. Java *sucks* at playing nice with other systems, unless you build up a "play nice with Java" layer in those other systems first. PHP, shitty as it is, is much better. Python and Perl are both much better (Python more than perl these days, imo) as glue code.
Java works best when it can run in long running, memory hogging, monolithic processes, where it's development and execution model can really shine (these are the circumstances you get those "faster than C" benchmarks). Java doesn't play nice with other systems, and interop via JNI can kill the GC optimizations that let it beat manual memory management for performance.
A cluster farm of commodity PCs is easily capable of handling 3000 transactions per day, assuming that your system isn't run by morons. I'd tend to attribute these problems to more general IT/software development issues, like the customers designing a more complex business process that doesn't address thier current problems, lack of adequate testing and customer feedback, important people (like end users) being left out of the design and testing phases and, of course, the ever popular "new shiny" syndrome.
Interestingly, Rythymbox crashes when importing my collection (I suspect its a couple files with Unicode characters in them), while Amarok handles it just fine. Amarok is also has far, far more functionality than RB, although I don't use too much of it.
It's absurd to claim that playing games doesn't affect behavior, but it's equally absurd to claim it *causes* behavior. Especially without any evidence, of which there isn't any.
And claiming that its any more or less "real" than movies is total bullshit - in fact, my children are far more likely to act out shows on TV than they are the games they play, because if they want to act out the game, they can simply *play the game*.
Your argument can be summed up very simply: blah blah, I can't imagine anything other than my opinion, blah, the government and the entire community should work to maintain my standards, oh, and all the stuff I like doesn't count. Thats a pretty lousy basis for a system of governing, but whatever floats your boat.
The fact is, recycling doesnt work. More energy is consumed "recylcing" stuff than it would take to make a new one.
Not always true, but the point is not pure conservation of energy, but conservation of a rare resource at the expense of a more prolific one. Not all energy is the same.
But besides that - recycling many things produces vastly inferior products to the original (particularly with metals).
And thats a really good reason to never recycle anything, is it?
And, finally (the trollish-sounding part of my post), some of us feel it is our right to destroy the environment. I know that I do my part to destroy it bit by bit. Why? No good reason, honestly. Though if I really wanted to stretch, I'd say to give humanity a good reason to get off this rock;)
I know that nihilistic self-destruction seems really cool when you're a teenager, but eventually you *do* need to grow up.
I'm expecting its interface to make as much sense as Windows' interface.
All else aside, this is where you went wrong. The Windows interface *doesn't* make sense, unless you already know it. You're confused and distracted by something unfamiliar when you want it to work like Windows - except you don't, because you don't *have* a Xandros Network on Windows.
People hold "the Linux desktop" to a far higher standard than they do Windows. I used Ubuntu at home and Windows at work, and I am *constantly* frustrated by crap Windows does stupidly. Printing to my network printer locks the desktop until its done spooling. I have to reboot every 3 days because of patches corporate IT pushes to me. I had to reboot twice after installing Visio - even when I explicitly didn't install any of the shared components, so I shouldn't have worried about other Office apps being open. Accessing shared drives locks the desktop when theres a network hiccup.
Installing the nvidia drivers on Linux is only very slightly more difficult than installing them on Windows, assuming that your distro doesn't package them for you - on Ubuntu, it's far easier.
The fact is, the Linux desktop experience is generally equal to the Windows one - Windows is better in some ways (it's easier to get SMB file sharing working), Linux better in others (SMB performance is far superior once it's up). The Linux desktop experience is certainly *much* better than many previous versions of Windows - better than Win95 in essentially every way, for example. And you can't reasonably argue that Win95 'wasn't ready for the desktop'.
This was obvious in 1994, too. It's an obvious solution given the problem. The problem is that the patent (similiar to Amazons 1-click) is that it's a patent for a solution to a problem that people at the time didn't feel was worth solving. It's not a matter of "this problem is hard, and people should be rewarded for the innovation of developing it", it's a matter of "this problem is very simple, but the solution will cost money and we don't feel its worth our time to develop". The damage these patents would have caused to our economy if they'd been issued 50 years ago would have been enormous, and the damage they do to our economy now will be. Different solutions to business problems should not be patentable.
Further, the fact that independent invention is not a defense against patent infringment is incredibly injust and damaging. And the fact that our patent system actively discourages patent research for ideas (because we allow very broad patents, because patents are not practically required to actually detail thier implementation, and because of the increased penalties and low barrier to proving "knowing infringment") is also damaging and contrary to the fundamental purpose of patents.
Lastly, this patent, like the Eolas patent, *doesn't include the hard stuff*. Sanely implementing a pay as you go system involves a lot of technologies besides a database lookup of the account number - thats the ludicrously obvious and simple part of it. Just as the Eolas patent covers the concept of plugins but doesn't provide any of the hard details, like implementing the object system/api you'll need for interaction.
Why is the guy at Intel an asshole if he solves the same problem you did? Personally, I think independent invention should be considered de facto evidence of obviousness, and should invalidate a patent unless the patent holder can demonstrate otherwise, with a fairly signifigent burden. Further, even if they do make that burden, so long as the defendant can prove true independent invention, the holder should be forced to license it at no cost to the independent inventor. And process patents (like this one, and the Eolas patent) that simply describe a problem domain without an implementation of a solution should be tossed without delay.
See, this is what happens when people don't think things through. You are 100% correct in your reading of the DMCA. But what you are forgetting, and what has *already happened* is that regardless of whether an individual copyright holder authorizes it or not, the tools required to do so are illegal. The DMCA is incredibly well-crafted legislation, if you look at it from the point of view of a media cartel - your rights to free uses aren't (technically) being removed or violated. But your practical ability to exercise those rights is severly curtailed by the DMCA, both through outright banning of the tools needed to exercise those rights (DeCSS, for example), and through the chilling effect on the research needed to do it.
With just the timestamp and the IP Address your PC used to communicate with the central server, you can be easily traced. It's easier if you are on broadband, slightly more difficult if you are on a service like AOL or MSN.
For what it's worth, AOL maintains extensive logs and readily cooperates with law enforcement. I suspect that MSN does as well. I briefly assisted in a fraud investigation (purchasing stuff via our website with stolen credit cards) and the perpetuator was dialing in from an AOL account. AOL was able to take the source IP address and a timestamp and provide his account and billing information, as well as the telephone number he called from.
Slashdot Mirror
I feel obliged to point out that in addition to being wrong about the motivations, he's also wrong that Tanenbaum thought that monolithic kernels were slow. Everyone knew that monolithic kernels were faster - fewer context switches, no message-passing overhead. His claim was (and is) that microkernels were better, not that they were faster. And the success of Linux really doesn't address the fundamental design and architetural differences between micro and macrokernels at all.
It's a contract in an abstract meaning of the word, like you might say "social contract", but in (US) legal terminology, the difference is important. And since the grandparent was asking about breach of contract damages, thats the usage I was addressing.
You're defining successful intelligent life as "life which would have colonized Earth" and noting that because it didn't happen, there is no successful life. A species that managed to colonize *any* other planets is going to be one much more advanced than ours, and a species that actually maintained some sort of interstellar society far, far so. If such a species existed, they would neccesarily need forms of communication and travel we aren't aware of and cannot detect, and yet they still could easily be sufficently distant from us to not have colonized Earth. Note that I'm not suggesting such a species exists, and that while they *could*, it's reasonable to work from the assumption that they don't. But going from "they probably don't" to "they must not" exist is silly. Think of it as religion if you want - at this point, it's a matter of faith or assumption, not of evidence. Certainly, if other intelligences do exist, they are irrelevent to us this point (barring sci-fi ideas that they know about Earth and are intentionally avoiding it for whatever reason).
If you want to learn how to *program*, then you need an itch you want to scratch. Taking a programming course, or reading a book and working the samples will teach you about programming, and maybe computer science (depending on the book), but won't teach you how to program. And learning about programming, or computer science, or computer hardware and engineering might be what you want - it's an interesting topic. But if what is bothering you is something like "this program sucks, I need something better" or "it'd be nice this program did this thing", or "I need a way to automate this because it's annoying", then you want to get a book on a appropriate language (I heavily recommend against Perl or C++, try Python or Ruby for local stuff) and start hacking on it until your problem goes away.
This is totally retarded, and the OPs "problem" is equally retarded. If you have a problem with "other people" modifying "your code", then you need to fix yourself, not your database.
It's copyright violation, not breach of contract (the GPL is not a contract). An amusing point, that I don't believe has ever been explored, is that many GPL violations are large scale, *commercial* violations - that is, they expose the violator to criminal prosecution as well as civil liability.
See http://tiny.seul.org/en/ for someone who's already done the work of stripping down Linux to run on an 8meg 486.
Presuming that by "russian piracy sites" you mean sites that offer downloads legal in Russia, then there is nothing illegal about that. It's like claiming that it should be illegal for a company to pay it's workers in China less than the US minimum wage, or for a US company with a factory in Indonesia to not maintain US environmental and work hazard standards.
It's not theft, but it's not okay (there are lot of things that aren't theft that are still both illlegal and wrong), and they should comply with the GPL, and people shouldn't download stuff illegally from P2P applications.
Note that none of that means that the music industry isn't stupid for understanding the competition they face from illegal downloads and addressing it reasonably (you can be within your legal rights but still be stupid), or that it's not legitimate to legally buy your music from overseas where it's cheaper.
This "argument" relies upon an *enormous* amount of assumptions, to the point where I question the intellecual honesty of making it. How about we just say that we know of no other intelligent life in the Universe, and have no way of knowing that they exist or not at this point, and leave it at that.
It's not just MS. And in any case, the whole point of trademark law and associated things is that it's not a winning proposition to rely on people to act correctly. If there were no legal consequences from using the OSI mark without permission, would you still dismiss the possibility offhand?
Whether you approve or not, OSI certainly is *a* authority on what "open source" means, and the OSI certified logo is respected for that reason. Maybe not by you, but you don't have to respect that if you don't want to. As for why someone might want to use the logo without approval, you might as well ask why someone would want to put the UL logo on thier product.
See http://www-128.ibm.com/developerworks/views/openso urce/projects.jsp for a list of some of the projects they're actively involved with. This list doesn't seem to include projects that they have released to the community, such as Cloudscape.
The trademark he's talking about is the OSI certified symbol, not just the phrase "open source".
The political aspects of needing to outsource operations for CYA are totally irrelevent to the technology involved - there are plenty of companies more than willing to build server farms for you. In fact, there are more of them than there are companies willing to manage your 100 million dollar mainframe for you.
Lastly, the amount of traffic implied by 3000 containers a day, while certainly enterprise class, doesn't come any where near the high end of other existing systems.
Many, many crucial business and government systems run on server clusters. It's not any sort of career suicide to do it.
This is exactly contrary to my experience. Java *sucks* at playing nice with other systems, unless you build up a "play nice with Java" layer in those other systems first. PHP, shitty as it is, is much better. Python and Perl are both much better (Python more than perl these days, imo) as glue code.
Java works best when it can run in long running, memory hogging, monolithic processes, where it's development and execution model can really shine (these are the circumstances you get those "faster than C" benchmarks). Java doesn't play nice with other systems, and interop via JNI can kill the GC optimizations that let it beat manual memory management for performance.
A cluster farm of commodity PCs is easily capable of handling 3000 transactions per day, assuming that your system isn't run by morons. I'd tend to attribute these problems to more general IT/software development issues, like the customers designing a more complex business process that doesn't address thier current problems, lack of adequate testing and customer feedback, important people (like end users) being left out of the design and testing phases and, of course, the ever popular "new shiny" syndrome.
Interestingly, Rythymbox crashes when importing my collection (I suspect its a couple files with Unicode characters in them), while Amarok handles it just fine. Amarok is also has far, far more functionality than RB, although I don't use too much of it.
And claiming that its any more or less "real" than movies is total bullshit - in fact, my children are far more likely to act out shows on TV than they are the games they play, because if they want to act out the game, they can simply *play the game*.
Your argument can be summed up very simply: blah blah, I can't imagine anything other than my opinion, blah, the government and the entire community should work to maintain my standards, oh, and all the stuff I like doesn't count. Thats a pretty lousy basis for a system of governing, but whatever floats your boat.
Not always true, but the point is not pure conservation of energy, but conservation of a rare resource at the expense of a more prolific one. Not all energy is the same.
But besides that - recycling many things produces vastly inferior products to the original (particularly with metals).
And thats a really good reason to never recycle anything, is it?
And, finally (the trollish-sounding part of my post), some of us feel it is our right to destroy the environment. I know that I do my part to destroy it bit by bit. Why? No good reason, honestly. Though if I really wanted to stretch, I'd say to give humanity a good reason to get off this rock ;)
I know that nihilistic self-destruction seems really cool when you're a teenager, but eventually you *do* need to grow up.
All else aside, this is where you went wrong. The Windows interface *doesn't* make sense, unless you already know it. You're confused and distracted by something unfamiliar when you want it to work like Windows - except you don't, because you don't *have* a Xandros Network on Windows.
People hold "the Linux desktop" to a far higher standard than they do Windows. I used Ubuntu at home and Windows at work, and I am *constantly* frustrated by crap Windows does stupidly. Printing to my network printer locks the desktop until its done spooling. I have to reboot every 3 days because of patches corporate IT pushes to me. I had to reboot twice after installing Visio - even when I explicitly didn't install any of the shared components, so I shouldn't have worried about other Office apps being open. Accessing shared drives locks the desktop when theres a network hiccup.
Installing the nvidia drivers on Linux is only very slightly more difficult than installing them on Windows, assuming that your distro doesn't package them for you - on Ubuntu, it's far easier.
The fact is, the Linux desktop experience is generally equal to the Windows one - Windows is better in some ways (it's easier to get SMB file sharing working), Linux better in others (SMB performance is far superior once it's up). The Linux desktop experience is certainly *much* better than many previous versions of Windows - better than Win95 in essentially every way, for example. And you can't reasonably argue that Win95 'wasn't ready for the desktop'.
Further, the fact that independent invention is not a defense against patent infringment is incredibly injust and damaging. And the fact that our patent system actively discourages patent research for ideas (because we allow very broad patents, because patents are not practically required to actually detail thier implementation, and because of the increased penalties and low barrier to proving "knowing infringment") is also damaging and contrary to the fundamental purpose of patents.
Lastly, this patent, like the Eolas patent, *doesn't include the hard stuff*. Sanely implementing a pay as you go system involves a lot of technologies besides a database lookup of the account number - thats the ludicrously obvious and simple part of it. Just as the Eolas patent covers the concept of plugins but doesn't provide any of the hard details, like implementing the object system/api you'll need for interaction.
Why is the guy at Intel an asshole if he solves the same problem you did? Personally, I think independent invention should be considered de facto evidence of obviousness, and should invalidate a patent unless the patent holder can demonstrate otherwise, with a fairly signifigent burden. Further, even if they do make that burden, so long as the defendant can prove true independent invention, the holder should be forced to license it at no cost to the independent inventor. And process patents (like this one, and the Eolas patent) that simply describe a problem domain without an implementation of a solution should be tossed without delay.
See, this is what happens when people don't think things through. You are 100% correct in your reading of the DMCA. But what you are forgetting, and what has *already happened* is that regardless of whether an individual copyright holder authorizes it or not, the tools required to do so are illegal. The DMCA is incredibly well-crafted legislation, if you look at it from the point of view of a media cartel - your rights to free uses aren't (technically) being removed or violated. But your practical ability to exercise those rights is severly curtailed by the DMCA, both through outright banning of the tools needed to exercise those rights (DeCSS, for example), and through the chilling effect on the research needed to do it.
For what it's worth, AOL maintains extensive logs and readily cooperates with law enforcement. I suspect that MSN does as well. I briefly assisted in a fraud investigation (purchasing stuff via our website with stolen credit cards) and the perpetuator was dialing in from an AOL account. AOL was able to take the source IP address and a timestamp and provide his account and billing information, as well as the telephone number he called from.