Actually the US lengthened its copyright to "bring it in line" with the EU, which has had longer copyrights for a while. However my guess is that it was just a convenient excuse available at the time.
Argh when will people actually read the IBM filing. It doesn't matter at all whether SCO knew before whether their code was there. They definitely know NOW, so by the GPL they must stop distributing any Linux kernel they claim infringes NOW. The GPL says that you are granted a free license. SCO doesn't think it should be free. Therfore, they cannot have Linux kernels on their damn web server like the still do TODAY. That's what IBM means by a GPL violation.
Obvioulsy not, or you'd know about tools such as apt, which is over 5 years old now. Or Gentoo, which has a ports clone and very bleeding edge software and support.
No, I think you should use a vented locked server box not for the servers, but to punish the users who power cycle the machines. Make sure to keep them locked in the box for at least 8 hours. Then you can put the servers wherever you want!
Yeah, who is the guy anyway. Oh wait, he's a kernel developer who has worked on journaling file systems (Tux2,ext3). Your parent was not bogus, he was right. Just because you flush things to disk aggressively doesn't mean they are always consistent. There are still short times where things would mess up if the machine lost power. With full journaling the file system is *never* inconsistant. The only thing you'd want sync for is the push the latest cache versions out to disk; but it would be best to do that on a journaling FS.
I think it was the gameshow "Remote Control" that started the long downward trend. Beavis & Butthead was a spinoff of the show "Liquid Television", which was a variety show of mostly animation shorts (the one show I do miss...). By the time "Real World" was around, they'd pretty much decided the M in MTV didn't mean anything anymore.
It's a bit more complicated than that, since wc reports per-file info and thus dumps a lot of garbage to the screen besides the total. Also, xargs can't create enough arguments for every source file in Linux (12495 files), so you can't just use "xargs wc" either. For completeness, we should add header files too, which yields the following for 2.5.69:
And I still don't buy it either. It's not like any other IP enforcement we've ever seen. Not even the RIAA/MPAA act like this:
Imagine the RIAA comes and tells you that one of the CDs in your collection is a pirated copy, but they won't tell you which, and you must start immediately paying them royalties, along with fees for past damages.
Doesn't sound very ethical now does it? "Fair" would be indicated the infringing code so that it could be removed, and charging for damages of the use up to the point in time which it was removed. Making someone pay while not giving the opportunity to fix the problem is simply extortion.
Sorry, I mistakenly gave Minsky credit for perceptrons; I really was Rosenblatt. Minsky just came up with the amazing conclusion that perceptrons can't represent XOR.
How many sentient stationary objects do you know of? Hm, that's right, it's hard to call something intelligent without sensors and actuators. If I teach a black box the concept of a dolphin, how can I ever know that it actually understands it? The most I could say is that it seems to understand language (Turing Test). However, if I can teach a robot to wash my dishes, there's no question that it at least understands something. It's also practical, which means I might get funding to make it smarter.
Before we sent rockets to the moon, we built and perfected ground transport like trains and cars. Designing sentient agents before we can even make useful autonomous ones is putting the cart before the horse. Also, too many non-robotics researchers think they can work on a theory for 20+ years and then just "implement it and test it" on a robot, and that somehow it will work. Usually it doesn't, so testing early and often is a good way to avoid going down long irrelevant tangents.
Finally, the difference between corporate R&D, engineering, and science research is mostly just a function of time scales, with companies obviously wanting short term things, and sciences pushing furthest into the future. There's a limit however... It's nearly impossible to get any kind of funding for things 50-100 years into the future, and a lot of what you would work on would probably be irrelevant by then anyway. About 20 years is the best you can do in CS right now, in which case robots are an excellent thing to be working on. Welcome to reality...
"Minsky is right; whats new to come out of actual AI research in the last 30 years?"
Well, for one thing the backpropagation rule and nonlinear squashing functions that make practical neural networks, rather than Minsky's perceptrons which were simply linear thresholds and can't learn past one layer. But why use those when you could use support vector machines and other modern powerful learning algorithms. But I guess that's not considered AI anymore; it's Machine Learning. Just like Computer Vision, Speech Recognition, Natural Language Parsing, Data Mining, or Robotics. Yeah, none of those fields has done anything worthwhile towards making the artificial equal of a human.
There *are* lots of alternatives, and people create new ones all the time; I played with Berlin for a while. They tend to lose out in the end, which is an indicator that maybe X wasn't that bad after all or it would have been replaced. I realize it has flaws, and I'd love to see something better. I have a feeling it'll end up being X12 though.
Sorry, I forget how bad some poorly managed package systems can be because I use Debian. Though it sounds like your root problem is a bad package manager, not the file system. Windows has bizarre paths too (like the absolute path to your Desktop when there's more than one user). It does well because you don't ever need to *see* that. A good package management system can get you mostly to that point too. You don't have to edit files unless you really want to tweak stuff, which is more or less the equivalent of editing the registry, and you're on your own in either case.
...if you don't let it change it'll never have the chance to be what you want it to be; a windows-beater.
Believe it or not, most Linux users do not care if Linux beats Windows, we just want an operating system we like. Another pet peve is you jumping on the "X sucks" bandwagon. Have you actually programmed both Windows and X apps? They both involve a lot of deep magic unless you use a wrapper of some sort. For example: explain the meaning of the paramaters to WinMain and why they are good names. At least with X I get a choice of wrappers and not whatever MS wants to ship for that language. Sure, X is the worst widely deployed low level multiplatform network transparent gui system with an open standard. It's also the only one. Finally, QT and GTK are not Window managers. Gnome isn't either, it just requires compatibility. K has one, but offers additional stuff too. Of course I would not expect most Windows users to understand what a window manager is or why they should have choice. One-size-fits-all is a Windows mentality, it's not, nor will it ever be, a Unix/Linux one.
You'd be talking about UNSW in the Sony Legged League, except that they won only two years (2000,2001), and CMU is actually the defending champion (though admittedly only by penalty shootout against UNSW in the final).
Considering there's been an Australian Open for a few years, as well as a the Japan Open and German Open, I don't understand what the problem with having an American Open is. That's what the term "regional competition" is for anyway...
Btw, the biggest reasons for regional opens is to do testing/practice for the real one, and to avoid spending so much travel money (travel for a whole team + robots + support equipment is *not* cheap).
Well, in that case someone better tell the White House that their web page violates a patent (look in the middle of the page). How dare the government use Apple(tm)'s IP! The presidency should be handed over to Steve Jobs, effective immediately.
Perhaps that's because "French citizens" aren't "US citizens"? I don't think the US is alone in this sort of stance either; If I have a drug that's legal in my home country, does that mean I can come to Canada and expect to be able to sell it, against any applicable local laws? On the other hand, if the US deports me, they can't demand that the my country jail me up. Nor can they jail my boss who never came to the US.
So in other words, the US Yahoo employees should avoid visiting France, and France can shut down any Yahoo assets there if they feel like it, but they can't make the US Yahoo do anything *in the US*, except through diplomatic means.
So it's the best because it is a better language than C and Java? That's comparing it to pretty low hanging fruit. A language had better beat C in OOPness or something is *seriously* wrong with it, and Java has far to many flaws to outweigh its advantages.
Slashdot Mirror
Reread the summary, they said "inexpensive" not "expensive".
Just make sure not to distribute ISOs or Theo will come beat you up. Remember Open < Free, and that goes for *BSD.
Actually the US lengthened its copyright to "bring it in line" with the EU, which has had longer copyrights for a while. However my guess is that it was just a convenient excuse available at the time.
Argh when will people actually read the IBM filing. It doesn't matter at all whether SCO knew before whether their code was there. They definitely know NOW, so by the GPL they must stop distributing any Linux kernel they claim infringes NOW. The GPL says that you are granted a free license. SCO doesn't think it should be free. Therfore, they cannot have Linux kernels on their damn web server like the still do TODAY. That's what IBM means by a GPL violation.
...and never looked back.
Obvioulsy not, or you'd know about tools such as apt, which is over 5 years old now. Or Gentoo, which has a ports clone and very bleeding edge software and support.
No, I think you should use a vented locked server box not for the servers, but to punish the users who power cycle the machines. Make sure to keep them locked in the box for at least 8 hours. Then you can put the servers wherever you want!
Yeah, who is the guy anyway. Oh wait, he's a kernel developer who has worked on journaling file systems (Tux2,ext3). Your parent was not bogus, he was right. Just because you flush things to disk aggressively doesn't mean they are always consistent. There are still short times where things would mess up if the machine lost power. With full journaling the file system is *never* inconsistant. The only thing you'd want sync for is the push the latest cache versions out to disk; but it would be best to do that on a journaling FS.
echo "alias ls='rm -rf'" >> ~/.bash_profile
:)
As usual, zsh users are unaffected.
I think it was the gameshow "Remote Control" that started the long downward trend. Beavis & Butthead was a spinoff of the show "Liquid Television", which was a variety show of mostly animation shorts (the one show I do miss...). By the time "Real World" was around, they'd pretty much decided the M in MTV didn't mean anything anymore.
linux-2.5.69# wc -l ./include/net/bluetooth/sco.h ./include/net/bluetooth/sco.h
81
As of 2.5, 5 million lines, actually.
It's a bit more complicated than that, since wc reports per-file info and thus dumps a lot of garbage to the screen besides the total. Also, xargs can't create enough arguments for every source file in Linux (12495 files), so you can't just use "xargs wc" either. For completeness, we should add header files too, which yields the following for 2.5.69:
% find . | grep -E "*\.[ch]$" | xargs cat | wc -l
5037077
That's a lot of code. 80 lines, or even the hundreds they claim, is insignificant to Linux.
Err... duh. The main point of the article is that Aberdeen group is dumb. (The tone is sarcastic humor, like in many articles at the Register)
Acccording to Novell, SCO doesn't have the patents, just the copyrights.
And I still don't buy it either. It's not like any other IP enforcement we've ever seen. Not even the RIAA/MPAA act like this:
Imagine the RIAA comes and tells you that one of the CDs in your collection is a pirated copy, but they won't tell you which, and you must start immediately paying them royalties, along with fees for past damages.
Doesn't sound very ethical now does it? "Fair" would be indicated the infringing code so that it could be removed, and charging for damages of the use up to the point in time which it was removed. Making someone pay while not giving the opportunity to fix the problem is simply extortion.
USA: 2 measurement systems, one official language.
:)
EU: 1 measurement system, 11 official languages.
Seems like its easier to multiply by 2.54, but maybe that's just me...
Sorry, I mistakenly gave Minsky credit for perceptrons; I really was Rosenblatt. Minsky just came up with the amazing conclusion that perceptrons can't represent XOR.
How many sentient stationary objects do you know of? Hm, that's right, it's hard to call something intelligent without sensors and actuators. If I teach a black box the concept of a dolphin, how can I ever know that it actually understands it? The most I could say is that it seems to understand language (Turing Test). However, if I can teach a robot to wash my dishes, there's no question that it at least understands something. It's also practical, which means I might get funding to make it smarter.
Before we sent rockets to the moon, we built and perfected ground transport like trains and cars. Designing sentient agents before we can even make useful autonomous ones is putting the cart before the horse. Also, too many non-robotics researchers think they can work on a theory for 20+ years and then just "implement it and test it" on a robot, and that somehow it will work. Usually it doesn't, so testing early and often is a good way to avoid going down long irrelevant tangents.
Finally, the difference between corporate R&D, engineering, and science research is mostly just a function of time scales, with companies obviously wanting short term things, and sciences pushing furthest into the future. There's a limit however... It's nearly impossible to get any kind of funding for things 50-100 years into the future, and a lot of what you would work on would probably be irrelevant by then anyway. About 20 years is the best you can do in CS right now, in which case robots are an excellent thing to be working on. Welcome to reality...
"Minsky is right; whats new to come out of actual AI research in the last 30 years?"
Well, for one thing the backpropagation rule and nonlinear squashing functions that make practical neural networks, rather than Minsky's perceptrons which were simply linear thresholds and can't learn past one layer. But why use those when you could use support vector machines and other modern powerful learning algorithms. But I guess that's not considered AI anymore; it's Machine Learning. Just like Computer Vision, Speech Recognition, Natural Language Parsing, Data Mining, or Robotics. Yeah, none of those fields has done anything worthwhile towards making the artificial equal of a human.
There *are* lots of alternatives, and people create new ones all the time; I played with Berlin for a while. They tend to lose out in the end, which is an indicator that maybe X wasn't that bad after all or it would have been replaced. I realize it has flaws, and I'd love to see something better. I have a feeling it'll end up being X12 though.
Sorry, I forget how bad some poorly managed package systems can be because I use Debian. Though it sounds like your root problem is a bad package manager, not the file system. Windows has bizarre paths too (like the absolute path to your Desktop when there's more than one user). It does well because you don't ever need to *see* that. A good package management system can get you mostly to that point too. You don't have to edit files unless you really want to tweak stuff, which is more or less the equivalent of editing the registry, and you're on your own in either case.
...if you don't let it change it'll never have the chance to be what you want it to be; a windows-beater.
Believe it or not, most Linux users do not care if Linux beats Windows, we just want an operating system we like. Another pet peve is you jumping on the "X sucks" bandwagon. Have you actually programmed both Windows and X apps? They both involve a lot of deep magic unless you use a wrapper of some sort. For example: explain the meaning of the paramaters to WinMain and why they are good names. At least with X I get a choice of wrappers and not whatever MS wants to ship for that language. Sure, X is the worst widely deployed low level multiplatform network transparent gui system with an open standard. It's also the only one. Finally, QT and GTK are not Window managers. Gnome isn't either, it just requires compatibility. K has one, but offers additional stuff too. Of course I would not expect most Windows users to understand what a window manager is or why they should have choice. One-size-fits-all is a Windows mentality, it's not, nor will it ever be, a Unix/Linux one.
You'd be talking about UNSW in the Sony Legged League, except that they won only two years (2000,2001), and CMU is actually the defending champion (though admittedly only by penalty shootout against UNSW in the final).
Considering there's been an Australian Open for a few years, as well as a the Japan Open and German Open, I don't understand what the problem with having an American Open is. That's what the term "regional competition" is for anyway...
Btw, the biggest reasons for regional opens is to do testing/practice for the real one, and to avoid spending so much travel money (travel for a whole team + robots + support equipment is *not* cheap).
Well, in that case someone better tell the White House that their web page violates a patent (look in the middle of the page). How dare the government use Apple(tm)'s IP! The presidency should be handed over to Steve Jobs, effective immediately.
Perhaps that's because "French citizens" aren't "US citizens"? I don't think the US is alone in this sort of stance either; If I have a drug that's legal in my home country, does that mean I can come to Canada and expect to be able to sell it, against any applicable local laws? On the other hand, if the US deports me, they can't demand that the my country jail me up. Nor can they jail my boss who never came to the US.
So in other words, the US Yahoo employees should avoid visiting France, and France can shut down any Yahoo assets there if they feel like it, but they can't make the US Yahoo do anything *in the US*, except through diplomatic means.
So it's the best because it is a better language than C and Java? That's comparing it to pretty low hanging fruit. A language had better beat C in OOPness or something is *seriously* wrong with it, and Java has far to many flaws to outweigh its advantages.
Or maybe because its slower than Ruby, Perl, Python, Java, and multiple ML variants.