My favorite P2P protocal is the Internet Protocal. If ISPs are going to block P2P, they should start with that one. All the other ones rely on it anyway.
Children will learn there are rules in society that must be obeyed if they wish to live in that society.
Don't confuse the rules of society with your obsessive parenting. Everyone has to deal with the former, whereas only your unfortunate children are forced to deal with the latter.
But if you just stop paying attention to them and giving them 100% free reign all the time, you end up with a messed up kid.
The evidence I've seen strongly implies that the damage caused by extremely overprotective parents is worse. At least the abandoned kids learn to solve their own problems and interact with other people without supervision.
You sound like you were a spoiled brat whose parents needed to give a serious attitude adjustment. I would've taken away your computer for a couple of weeks if you spoke to me like that (or if you bypassed my measures), probably along with your cellphone, your ipod and all your music. And if you still had a bad attitude, I'd take your door off the hinges. If you STILL didn't get it, I'd come to school with you and follow you around, making sure your friends saw you, until you begged for mercy.
That is a battle a parent can *never* win against a determined teenager. This is an intelligent human being who lives in your house. Once you take enough away from them that they no longer think they have anything left to lose they'll either completely ignore you or start retaliating.
A much more reasonable solution is to realize that children are human beings and once they start valuing privacy and autonomy it's 100% reasonable to let them have some. Just because you're their parent doesn't mean you need to be administrating their life. Further, if you can't convince them to do what you want there's very little chance you'll productively force them.
Engine developers are very different from game developers. The option of doing Linux is a useful marketing bullet item. Having the option to target Linux if and when it becomes financially justifiable is nice. However offering such an option does not imply that native Linux versions are justifiable today. Keep in mind that engine developers are selling to people who won't have a finished game for years.
Sure, that doesn't change the fact that Valve *is* an engine developer and that Half Life is just as much an engine demo as Doom III or Unreal Tournament 2004 were. My point is simply this: Using the proposed logic (considering what other similar developers do), it isn't obvious that Valve shouldn't port to Linux.
I agree that cross platform development can help, however it makes far more sense to target Mac than Linux. Adding a third platform, Linux, would not improve things much over two, Win32 and Mac. Even when portions of a game are ported to Linux in order to create a server there is still a lot of work to be done with respect to getting the user interface and other client side code running.
Is it really the third platform, or the fifth? I hear that the PS3 and XBox360 are pretty big gaming platforms. When you look at it that way, you quickly realize that the Mac, Linux, and PS3 ports are largely the same code (OpenGL renderer) - as are the Windows and XBox360 ports (DirectX). At that point, the question of a release on Mac or Linux is basically installer testing rather than any sort of significant extra programming effort.
That seems to be why both ID and Epic make both Mac and Linux releases of their games. They're already writing a cross-platform game with an OpenGL renderer - releasing for a couple of extra similar platforms has trivial costs compared to a non-zero number of extra sales and some good PR. It also future-proofs their engines in case Linux happens to hit an inflection point in uptake during that engine's useful lifespan.
Now they're adding buggy compiz _by default_ What was wrong with that 1 click enable system they previously had?
My guess is that they'll be using a reasonably stable version of compiz. What's wrong with the one click disable system they'll have, anyway?
As for Dapper being more stable than the releases since, that's by design. The next extra-stable release (version 8.4, code name "Hardy Heron") will come out next April.
Just because many of the people who post on the Ubuntu forums are noobs doesn't mean that the distro itself is only for noobs. Sure, Ubuntu is just Debian with lazy defaults - but for many applications that happens to be exactly what is needed.
I just want to offload some X processing from my CPU to my graphics chip, to make the workstation run faster overall.
Compiz doesn't actually do that in practice yet. On a reasonably modern CPU + GPU combo it won't slow you down, but the current version doesn't speed you up compared to a traditional window manager. Further, your graphics card (GeForce 2 Go) is old enough that it would have to offload a significant chunk of the rendering work to the CPU anyway.
Based on the last benchmarks I saw, the minimum graphics card to not get a blatant performance loss from compiz with any useful effects is a mid-range card from 2003 or so (the GeForce 2 is from 2000).
Now you really _will_ be getting less by doing that - even the higher quality videos available over the internet fall far below the quality of DVB channels.
I'm pretty sure that an HR.HDTV pirate XviD compares favorably in quality to a digital signal downrezzed to NTSC.
Okay, I'm being pessimistic - but something pretty similar happened after Matrox released the specifications for its 3D graphics cards. There were fully open-source drivers, but they weren't exactly high-quality. I moved on to Nvidia after that...
Was Matrox even producing products at that point, or were you expecting one of the other six guys with old Matrox cards to support your drivers?
Firefox is still going strong, but Thunderbird lost its financial backing, the calender apps never make it to a 1.0 release, and the general opinion in my shop is that we need to go back to Outlook.
Wait a second... you switched to a piece of software that didn't meet your requirements, and now you're complaining that it didn't magically gain the features you wanted? Your problem there isn't open source software or commercial success - it's simply mis-evaluating a piece of software.
For 90% of users, programs are a tool to get work done, not something that you fiddle with and discuss the merit of licensing philosophies over.
Looking around you, I'm sure you can see that a number of people think that software licensing is a practically relevant issue. The questions you should be asking yourself is this: Why do they think it's important? Could there be directly practical results of software licensing that you are seeing because of the way you're looking at the issue?
While Stallman would like to bully companies into throwing open their source on anything that might touch his precious GNU project, its not realistic. If you want companies to write drivers for their hardware and ensure compatability, you have to give them the option to keep their secrets secret. Otherwise you'll just scare them off and hurt yourself in the process.
That's an interesting claim, but it happens to be false.
Most hardware companies sell hardware and want people to buy it. That means that they want their customers to be able to use it. Providing specifications (and editable driver code) makes the hardware maximally useful to the customer - thus selling the most units of the hardware. Usually the hardware interfaces are actually standardized, so keeping them secret is absurd.
You can see this in practice by looking at what hardware is supported by the Linux kernel today. Almost all hardware is supported by drivers embedded in the kernel itself. The only significant exceptions to that are high end 3D video cards, wireless network cards, and software modems. For 3D cards, one of the two vendors in the world just came to their senses and realized that hiding interface details just costs them sales for no good reason. For wireless cards and software modems, the excuse is legal constraints.
That leaves nVidia as the only company in the entire world that may legitimately be thinking like you are. My guess is that they aren't thinking like that - they're probably just being lazy and don't want to pay their lawyers to evaluate the exact legal situation of the various patent licenses that they have that may restrict what information they can publicly release.
That's why Linus' productive output surpasses Stallman's by a factor 1000.
The hell? Not only is that irrelevant, it's also blatantly wrong. If you try to build a GNU/Linux system without Stallman's code you'll have just as much trouble as if you tried to do it without Linus's code - and the number of lines of involved code actually written by each is probably pretty similar.
Maybe at first he was on that mission, but as far as I can tell from recent comments by RMS, his mission has become to promote Richard Stallman and the "Only my freedom is really free enough" point of view.
Richard Stallman has been giving the same answers to the same questions for 10+ years now. His opinions are well considered - you should take the time to understand his positions and why he holds them. Once you do that, we can have a rational discussion about ideas rather than a completely irrelevant discussion about a geek not being an eloquent public speaker from the perspective of people who don't already understand accept his reasoning behind radical-sounding opinions.
Everyone has their own "expert opinion" on what constitutes "good code."
And apparently yours is "not Perl or C++", even though those two languages have been used to build useful real-world systems with some regularity for years and there are accepted community standards for writing good code in both. I think Bjarne Stroustrup said it best: "There are just two kinds of languages: the ones everybody complains about and the ones nobody uses."
Object orientation in PERL is not part of the core language, but is a hack based on the way that the symbol table works. You can not possibly seriously suggest that it's a good idea.
Why not? Writing object oriented code in Perl works great in practice. Just because the syntax extensions to allow it didn't need to be that drastic doesn't mean the result isn't good.
if open office has a feature that word also has, open office gets declared better I don't know why.
Have you considered that the author of the article may have compared the implementation of the features and used that as the basis for his judgment?
Sure, rarely used unique features are neat - but it's the usability of common features that matters most for the usability of something as well-defined as a word processor.
Anyone who thinks they can write substantial programs (>20 lines (yes, I am being sarcastic)) in PERL that will be maintainable, reliable and testable is severely misguided.
Anyone who doesn't think this is possible has never learned Perl. Perl's greatest weakness is that it looks just enough like other languages that people expect to be able to just jump in and read / write Perl code without actually learning Perl - and then they hit some common perl idiom like an inside-out class (or even just the code to access a nested data structure), have no idea what's going on, and blame the language.
Other than building cellulostic ethanol factories, and some ethanol pipelines, we alredy have everything else (unlike corn, sugarbeets, biodiesel, hydrogen, dirtect electric, or other proposed systems).
Why are any of the biodiesel plans (say switchgrass to biodiesel) any different from that? Hell, they're a bit better because there are already a significant number of diesel vehicles and there is already a complete diesel fuel distribution system in place.
Dude! Where can I get these diesel engines with 6 times the fuel economy of my gasoline car? (By the way, my car gets about 36 mpg - gasoline - on the highway...)
He was exaggerating in one direction and you are intentionally mis-interpreting it in the other direction. A six-fold decrease in gallons of fuel consumed overall if the average American car was replaced with an "efficient diesel" (that would be used for similar tasks in Europe or Asia - so an SUV becomes a VW Golf Diesel) is actually pretty reasonable.
So the people that want to use the code pay the people that wrote it (and hold the copyright) for the privilege of licensing it under a BSD license.
Close, but wrong in a very important way. It's never "pay for a BSD license". Rather, it's "pay for a proprietary license". A BSD license would allow the recipient to redistribute the code under the BSD license (which would completely break the GPL licensing plan).
Slashdot Mirror
My favorite P2P protocal is the Internet Protocal. If ISPs are going to block P2P, they should start with that one. All the other ones rely on it anyway.
Oh really?
What other device is a wireless node and acts like a wireless mesh router even when powered off?
What other device has a 1200x900 screen that takes well under a watt?
Sure, it's not a huge leap ahead of other, similar devices - but the XO is definitely pushing the boundries of mobile computer design.
I'm pretty sure that the EFF is there to promote *liberties* not *anti-community-restrictions* like the CC NC clause.
Don't confuse the rules of society with your obsessive parenting. Everyone has to deal with the former, whereas only your unfortunate children are forced to deal with the latter.
The evidence I've seen strongly implies that the damage caused by extremely overprotective parents is worse. At least the abandoned kids learn to solve their own problems and interact with other people without supervision.
That is a battle a parent can *never* win against a determined teenager. This is an intelligent human being who lives in your house. Once you take enough away from them that they no longer think they have anything left to lose they'll either completely ignore you or start retaliating.
A much more reasonable solution is to realize that children are human beings and once they start valuing privacy and autonomy it's 100% reasonable to let them have some. Just because you're their parent doesn't mean you need to be administrating their life. Further, if you can't convince them to do what you want there's very little chance you'll productively force them.
Sure, that doesn't change the fact that Valve *is* an engine developer and that Half Life is just as much an engine demo as Doom III or Unreal Tournament 2004 were. My point is simply this: Using the proposed logic (considering what other similar developers do), it isn't obvious that Valve shouldn't port to Linux.
Is it really the third platform, or the fifth? I hear that the PS3 and XBox360 are pretty big gaming platforms. When you look at it that way, you quickly realize that the Mac, Linux, and PS3 ports are largely the same code (OpenGL renderer) - as are the Windows and XBox360 ports (DirectX). At that point, the question of a release on Mac or Linux is basically installer testing rather than any sort of significant extra programming effort.
That seems to be why both ID and Epic make both Mac and Linux releases of their games. They're already writing a cross-platform game with an OpenGL renderer - releasing for a couple of extra similar platforms has trivial costs compared to a non-zero number of extra sales and some good PR. It also future-proofs their engines in case Linux happens to hit an inflection point in uptake during that engine's useful lifespan.
You do realize that Valve is the *only* major game engine vendor that still doesn't provide a native Linux port, right?
My guess is that they'll be using a reasonably stable version of compiz. What's wrong with the one click disable system they'll have, anyway?
As for Dapper being more stable than the releases since, that's by design. The next extra-stable release (version 8.4, code name "Hardy Heron") will come out next April.
Just because many of the people who post on the Ubuntu forums are noobs doesn't mean that the distro itself is only for noobs. Sure, Ubuntu is just Debian with lazy defaults - but for many applications that happens to be exactly what is needed.
Compiz doesn't actually do that in practice yet. On a reasonably modern CPU + GPU combo it won't slow you down, but the current version doesn't speed you up compared to a traditional window manager. Further, your graphics card (GeForce 2 Go) is old enough that it would have to offload a significant chunk of the rendering work to the CPU anyway.
Based on the last benchmarks I saw, the minimum graphics card to not get a blatant performance loss from compiz with any useful effects is a mid-range card from 2003 or so (the GeForce 2 is from 2000).
I'm pretty sure that an HR.HDTV pirate XviD compares favorably in quality to a digital signal downrezzed to NTSC.
Was Matrox even producing products at that point, or were you expecting one of the other six guys with old Matrox cards to support your drivers?
Wait a second... you switched to a piece of software that didn't meet your requirements, and now you're complaining that it didn't magically gain the features you wanted? Your problem there isn't open source software or commercial success - it's simply mis-evaluating a piece of software.
Looking around you, I'm sure you can see that a number of people think that software licensing is a practically relevant issue. The questions you should be asking yourself is this: Why do they think it's important? Could there be directly practical results of software licensing that you are seeing because of the way you're looking at the issue?
That's an interesting claim, but it happens to be false.
Most hardware companies sell hardware and want people to buy it. That means that they want their customers to be able to use it. Providing specifications (and editable driver code) makes the hardware maximally useful to the customer - thus selling the most units of the hardware. Usually the hardware interfaces are actually standardized, so keeping them secret is absurd.
You can see this in practice by looking at what hardware is supported by the Linux kernel today. Almost all hardware is supported by drivers embedded in the kernel itself. The only significant exceptions to that are high end 3D video cards, wireless network cards, and software modems. For 3D cards, one of the two vendors in the world just came to their senses and realized that hiding interface details just costs them sales for no good reason. For wireless cards and software modems, the excuse is legal constraints.
That leaves nVidia as the only company in the entire world that may legitimately be thinking like you are. My guess is that they aren't thinking like that - they're probably just being lazy and don't want to pay their lawyers to evaluate the exact legal situation of the various patent licenses that they have that may restrict what information they can publicly release.
The hell? Not only is that irrelevant, it's also blatantly wrong. If you try to build a GNU/Linux system without Stallman's code you'll have just as much trouble as if you tried to do it without Linus's code - and the number of lines of involved code actually written by each is probably pretty similar.
Richard Stallman has been giving the same answers to the same questions for 10+ years now. His opinions are well considered - you should take the time to understand his positions and why he holds them. Once you do that, we can have a rational discussion about ideas rather than a completely irrelevant discussion about a geek not being an eloquent public speaker from the perspective of people who don't already understand accept his reasoning behind radical-sounding opinions.
And apparently yours is "not Perl or C++", even though those two languages have been used to build useful real-world systems with some regularity for years and there are accepted community standards for writing good code in both. I think Bjarne Stroustrup said it best: "There are just two kinds of languages: the ones everybody complains about and the ones nobody uses."
Why not? Writing object oriented code in Perl works great in practice. Just because the syntax extensions to allow it didn't need to be that drastic doesn't mean the result isn't good.
Have you considered that the author of the article may have compared the implementation of the features and used that as the basis for his judgment?
Sure, rarely used unique features are neat - but it's the usability of common features that matters most for the usability of something as well-defined as a word processor.
Anyone who doesn't think this is possible has never learned Perl. Perl's greatest weakness is that it looks just enough like other languages that people expect to be able to just jump in and read / write Perl code without actually learning Perl - and then they hit some common perl idiom like an inside-out class (or even just the code to access a nested data structure), have no idea what's going on, and blame the language.
Why are any of the biodiesel plans (say switchgrass to biodiesel) any different from that? Hell, they're a bit better because there are already a significant number of diesel vehicles and there is already a complete diesel fuel distribution system in place.
He was exaggerating in one direction and you are intentionally mis-interpreting it in the other direction. A six-fold decrease in gallons of fuel consumed overall if the average American car was replaced with an "efficient diesel" (that would be used for similar tasks in Europe or Asia - so an SUV becomes a VW Golf Diesel) is actually pretty reasonable.
Close, but wrong in a very important way. It's never "pay for a BSD license". Rather, it's "pay for a proprietary license". A BSD license would allow the recipient to redistribute the code under the BSD license (which would completely break the GPL licensing plan).
As legal documents go, the GPL isn't that bad. Hell, the license itself is shorter than the GPL FAQ you quoted.
In any case, if you haven't read the GPL and understood it's mechanism then you probably shouldn't be trying to tell people what it does.