Would you mind posting or providing a link to information about your design, we can benefit from it too?
I, for one, would like to build an ARM PDA with buttons along the sides which I can chord with to avoid the touchpad/handwriting recognition/etc. problem.
But I don't know much about motherboard design, so looking at someone else's work would be helpful.
Mostly I was just saying that your analogy isn't valid; you can't directly equate information with physical items until we have replicators and free energy.
But even in a world without copyright, labor would still be compensated. The proof of this is obvious: before copyright existed people were paid for their labor. People even wrote, composed, and painted some of the most impressive art ever created. They were paid by commisions, were born into wealthy families, or did non-artistic jobs on the side.
Programmers have little to fear from copyright disappearing; there will always be the need for custom versions of programs and support. Neither do architects; many people want buildings for specific purposes or just want their house to be different. Writers, musicians, and other artists have been funded by governments and large organizations throughout history, and today too. Even without this, responding to desire for specific individual products should be enough to let most of them earn a living.
Perhaps people would stop becoming millionares with "intellectual property," but that wouldn't be such a terrible thing.
The difference is that with a car, labor has to be performed with every car made, whereas with online music distribution, the marginal cost is zero.
A proper analogy would be looking at your neighbor's car (with his permission) and building your own car, modeling it after his. The car company can still sue you for doing this, but it seems a lot less like theft to most people.
You are completely out of your mind if somebody is going to invest $200 million in something which they are going to give away.
While they might not be able to recoup $200 million on a single movie, perhaps they could recoup $10 million without copyright, by having movies be comissioned. This would probably reduce the number of movies being made and make actors only well-to-do instead of megawealthy, but I would find this an acceptable trade so we could reduce the limits of freedom of speech.
Consider the encrypted parts of Windows 3 beta which made Windows give error messages under DR-DOS. This exception would allow the creators of DR-DOS to reverse engineer Windows to allow DR-DOS not to cause the warnings.
However, if Microsoft had been more ruthless (and DMCA had been passed 12 years ago), and had made all programs running under Windows encrypted, then it would be a case of access control; only authorized (by the copyright holders of the programs) versions of Windows would be allowed to run Windows programs.
Note that you can still copy these hypothetical windows programs; you just can't run them. The DMCA does not prohibit circumventing copyright protection. It prohibits circumventing access protection. There is a subtle difference.
The judge said that he was uncertain code should be considered free speech since it is mostly utilitarian instead of expressive. But, he said, for the purposes of the injunction, he assumed that it was covered by free speech, and made his ruling with that in mind.
He argues that even if it is free speech, congress is still allowed to regulate it as they do other protected speech for economic reasons, as in copyright law.
I disagree with him that prohibiting programs which mainly let users exercise their fair use rights has anything to do with even indirectly promoting the useful arts and science. However, the judge is still open to criticizing the law for its constitutionality, though he cites the precedent that the courts should give leeway to congress in matters of applying laws to new technologies.
It said 3 windows per day. Assuming these are evenly spaced, that would be 24 / 3 = 8 hours per window. So perhaps they could work three times as fast with more processing power. Not a huge jump and the effort to get this going distributed would probably be more effort than its worth, though.
I can understand the press leaving out the 'GNU' when discussing Redhat or such. But Debian is very good at using the words GNU and Linux consistently; you'd hope it'd rub off on people who talk about them. Linus wrote Linux, a clone of the unix kernel. Debian is a GNU/Linux distribution. Oh well.
: Actually yes I do. Prevasive multithreading is NOT only advantagous : for SMP machines. the PII has 3 integer units and multithreaded : programs allow them to be use more effiently.
Obviously you don't know what an integer unit is or what multithreading means. The multiple integer units mean that the PII is a super-scalar processor. These units only run consecutive integer instructions. Multithreading on a single-processor machine means you switch between tasks on a single processor, introducing a context-switch, which multiple integer units don't help at all.
: Second, Be has test where BeOS pushes 600 frames of 2D per second using a dual PII machine and BDirectWindow.
I'm sure svgalib could do the same, but it's irrelevant as your monitor refresh is much less than 600 fps.
: It is much lighter weight
Linux is quite definitely lighter weight than BeOS as linux can be run without a GUI.
I'm glad 'Open Source' can't be OSI's certification mark, as 'Open Source' is a misnomer. If something is open and you're allowed near it, you can see inside it. Can you play with it? Perhaps there isn't a wall between you and its insides, but there may still be an agreement that you're not allowed to touch. Now OSI will have to be honest and say that a license is 'OSI Certified' which is what the license is if they okay it. They're not (or at least they shouldn't be) claiming their approval means the approval of the community. But I'd rather have companies use a (L)GPL, BSD, MIT, or X license rather than invent some new license with new restrictions to figure out how to abide by. Interopability of code is nice and a dozen new incompatible 'OSI Certified' licenses doesn't seem so nice. The problem is companies are used to the way they've been doing things and instead of being shown how we do things, they're being shown hype and, if we're lucky, the products of our doing. Are we a culture of technical and marketing achievements, or of sharing?
If you want to store information to send to the ICQ server, and you need the cleartext to authenticate, there's no good way. Anything you do can be undone, so you're not really providing actual security by obfuscating the stored password. This doesn't apply soley to visible/open source applications, either, as programs can be disassembled, or the program can be set to authenticate to a different server which listens to passwords, etc.
Password security relies on having a secret, and unless there's something the computer doesn't know, there is no secret, and the original password is retrievable. The MAC address isn't a secret.
If you obfuscate the password, perhaps you're just providing a false sense of security for the user, and you shouldn't do that. At least tell them it's not well protected. Rely on file permissions to keep other users from seeing their password.
But obfuscation can prevent accidental viewing from people who aren't actually after the password, but might be going through your files (backup administrators or something). Something simple like xoring the password with 0xAA is good.
If you want security, require the user to enter a password. If the password isn't terribly important (as in this case), don't worry about it.
If you want the analogy to be cars, then Linux is definitely the engine, and GNU the rest of car, the tools, and the factories that manufacture it all.
Do you refer to your hot rod by the name of its engine? Well, it's an important part of your car, so maybe?
Err, it's not just the libraries that are multi-threaded, it's the whole OS ! Kernel, filesystems, GUI, etc... there are threads everywhere, even in drivers. Linux is far from that right now.
Linux has threads in the kernel, GUI, and various programs. Specifically where is it lacking support? Yes, there is room for improvement, but I can't think of any big architectural changes necessary for those improvements.
Linux does SMP too, but that doesn't mean it really takes advantage of it. You need lots of multithreading in the OS itself to really use SMP. The "SMP capable" doesn't mean "SMP efficient". On Linux you need to thread your app to take advantage of SMP. On BeOS even a non threaded app will take advantage of it because all the system calls (kernel or GUI) go to multithread code.
Indeed, you need to thread your app to make it SMP efficient as opposed to SMP capable. Unless your application spends all its time in parallelizable system calls, SMP won't help it much unless it's threaded. If you don't write your application to be threaded, you're not going to take much advantage of SMP. Good kernel SMP support takes a lot of time and resources, neither of which Be has had. It's amazing they've done as well as they have.
Linux is too command-line oriented
For me, a command line is, for simple tasks, about three times faster than a GUI, and several hundred times faster for complex tasks. However, I can run GNU/Linux without ever touching a command-line if I want. And yes, BeOS can run a command line too, so I don't see what this statement is supposed to establish.
when Unix was created graphic display was sci-fi, and when XWindows was created digital video was sci-fi too. A real multimedia OS must be built from the start to fit into today hardware (accelerated 2D, 3D, video, etc..), and have all sort of stuff to synchronize and mix audio and video. Linux is rather poor here.
"When BeOS was created, holographic displays were sci-fi." I suppose when the next newest technology comes out you'll feel the need to write another OS? I'd rather take a solid foundation and build upon it. Linux intentionally doesn't have much graphics support; they decided to put as much as possible in userland, but X that doesn't mean the OS can't synchronize and mix audio and video. Perhaps you can point at some numbers? And don't just say 'BeOS can play eight videos at the same time' because I can do that too with a lowly K6-233 and a crappy video card.
Not to say that Linux is bad, it's an excellent server OS, makes Apache and Samba rocks, but it's a very poor thing when video and audio. It was just never design to handle that. You can't just add stuff here and there to make it more "multimedia", it just makes the OS more bloated and inefficient.
Actually, adding it on modularly makes it less bloated, and the efficiency cost isn't that terrible.
apt-get update; apt-get dist-upgrade
on
RedHat 6.0 is Out
·
· Score: 1
You mean you have to reboot to upgrade Redhat? Bah, I think I'll stick to Debian.
You have to edit a few lines in the kernel to get it to support more than 960 MB of physical memory. What I found interesting was that they apparently didn't make a separate swap partition for the linux box (they said 1 OS partition and 1 data partition)... hm...
: I don't think source code is any more a form of speech than a recipe : for a cake, a set of instructions to make a bomb, or a to-do list to : get one through a day.
By showing that the phrase 'open source' simply means 'open source' perhaps ESR will lose his battle to popularize Free software by hiding the freedom associated with it.
There's nothing ESR can do to stop their usage of the term, and in this case, Microsoft is actually correct.
You don't understand. Linux is a kernel. RMS does *not* want to rename Linux. He considers it the authors' right to call it whatever they want. GNU is an idea, an operating system, a collection of libraries and utilities. He wants to call the combination of GNU with Linux "GNU/Linux".
The answer to this question is obvious: Jon likes SexBots.
Would you mind posting or providing a link to information about your design, we can benefit from it too?
I, for one, would like to build an ARM PDA with buttons along the sides which I can chord with to avoid the touchpad/handwriting recognition/etc. problem.
But I don't know much about motherboard design, so looking at someone else's work would be helpful.
Mostly I was just saying that your analogy isn't valid; you can't directly equate information with physical items until we have replicators and free energy.
But even in a world without copyright, labor would still be compensated. The proof of this is obvious: before copyright existed people were paid for their labor. People even wrote, composed, and painted some of the most impressive art ever created. They were paid by commisions, were born into wealthy families, or did non-artistic jobs on the side.
Programmers have little to fear from copyright disappearing; there will always be the need for custom versions of programs and support. Neither do architects; many people want buildings for specific purposes or just want their house to be different. Writers, musicians, and other artists have been funded by governments and large organizations throughout history, and today too. Even without this, responding to desire for specific individual products should be enough to let most of them earn a living.
Perhaps people would stop becoming millionares with "intellectual property," but that wouldn't be such a terrible thing.
The difference is that with a car, labor has to be performed with every car made, whereas with online music distribution, the marginal cost is zero.
A proper analogy would be looking at your neighbor's car (with his permission) and building your own car, modeling it after his. The car company can still sue you for doing this, but it seems a lot less like theft to most people.
While they might not be able to recoup $200 million on a single movie, perhaps they could recoup $10 million without copyright, by having movies be comissioned. This would probably reduce the number of movies being made and make actors only well-to-do instead of megawealthy, but I would find this an acceptable trade so we could reduce the limits of freedom of speech.
Consider the encrypted parts of Windows 3 beta which made Windows give error messages under DR-DOS. This exception would allow the creators of DR-DOS to reverse engineer Windows to allow DR-DOS not to cause the warnings.
However, if Microsoft had been more ruthless (and DMCA had been passed 12 years ago), and had made all programs running under Windows encrypted, then it would be a case of access control; only authorized (by the copyright holders of the programs) versions of Windows would be allowed to run Windows programs.
Note that you can still copy these hypothetical windows programs; you just can't run them. The DMCA does not prohibit circumventing copyright protection. It prohibits circumventing access protection. There is a subtle difference.
The judge said that he was uncertain code should be considered free speech since it is mostly utilitarian instead of expressive. But, he said, for the purposes of the injunction, he assumed that it was covered by free speech, and made his ruling with that in mind.
He argues that even if it is free speech, congress is still allowed to regulate it as they do other protected speech for economic reasons, as in copyright law.
I disagree with him that prohibiting programs which mainly let users exercise their fair use rights has anything to do with even indirectly promoting the useful arts and science. However, the judge is still open to criticizing the law for its constitutionality, though he cites the precedent that the courts should give leeway to congress in matters of applying laws to new technologies.
It said 3 windows per day. Assuming these are evenly spaced, that would be 24 / 3 = 8 hours per window. So perhaps they could work three times as fast with more processing power. Not a huge jump and the effort to get this going distributed would probably be more effort than its worth, though.
I can understand the press leaving out the 'GNU' when discussing Redhat or such. But Debian is very good at using the words GNU and Linux consistently; you'd hope it'd rub off on people who talk about them.
Linus wrote Linux, a clone of the unix kernel. Debian is a GNU/Linux distribution. Oh well.
: Actually yes I do. Prevasive multithreading is NOT only advantagous
: for SMP machines. the PII has 3 integer units and multithreaded
: programs allow them to be use more effiently.
Obviously you don't know what an integer unit is or what multithreading means. The multiple integer units mean that the PII is a super-scalar processor. These units only run consecutive integer instructions. Multithreading on a single-processor machine means you switch between tasks on a single processor, introducing a context-switch, which multiple integer units don't help at all.
: Second, Be has test where BeOS pushes 600 frames of 2D per second using a dual PII machine and BDirectWindow.
I'm sure svgalib could do the same, but it's irrelevant as your monitor refresh is much less than 600 fps.
: It is much lighter weight
Linux is quite definitely lighter weight than BeOS as linux can be run without a GUI.
[ more drivel snipped ]
I'm glad 'Open Source' can't be OSI's certification mark, as 'Open Source' is a misnomer. If something is open and you're allowed near it, you can see inside it. Can you play with it? Perhaps there isn't a wall between you and its insides, but there may still be an agreement that you're not allowed to touch.
Now OSI will have to be honest and say that a license is 'OSI Certified' which is what the license is if they okay it. They're not (or at least they shouldn't be) claiming their approval means the approval of the community.
But I'd rather have companies use a (L)GPL, BSD, MIT, or X license rather than invent some new license with new restrictions to figure out how to abide by. Interopability of code is nice and a dozen new incompatible 'OSI Certified' licenses doesn't seem so nice.
The problem is companies are used to the way they've been doing things and instead of being shown how we do things, they're being shown hype and, if we're lucky, the products of our doing. Are we a culture of technical and marketing achievements, or of sharing?
And I suppose you say "I drove to work this morning in my spark plugs" when you used your entire car to get there?
Linux is a kernel, a small but important part of the GNU/Linux OS.
libc5 is based on gnu libc 1
Like yours?
If you want to store information to send to the ICQ server, and you need the cleartext to authenticate, there's no good way. Anything you do can be undone, so you're not really providing actual security by obfuscating the stored password. This doesn't apply soley to visible/open source applications, either, as programs can be disassembled, or the program can be set to authenticate to a different server which listens to passwords, etc.
Password security relies on having a secret, and unless there's something the computer doesn't know, there is no secret, and the original password is retrievable. The MAC address isn't a secret.
If you obfuscate the password, perhaps you're just providing a false sense of security for the user, and you shouldn't do that. At least tell them it's not well protected. Rely on file permissions to keep other users from seeing their password.
But obfuscation can prevent accidental viewing from people who aren't actually after the password, but might be going through your files (backup administrators or something). Something simple like xoring the password with 0xAA is good.
If you want security, require the user to enter a password. If the password isn't terribly important (as in this case), don't worry about it.
If you want the analogy to be cars, then Linux is definitely the engine, and GNU the rest of car, the tools, and the factories that manufacture it all.
Do you refer to your hot rod by the name of its engine? Well, it's an important part of your car, so maybe?
Um, no. RMS calls the kernel Linux, and the entire system GNU/Linux.
Meept or somebody calls it LiGnuX.
And because of BSD/Linux and GNU/Hurd, we may soon have practical reasons to call it GNU/Linux.
Err, it's not just the libraries that are multi-threaded, it's the whole OS ! Kernel, filesystems, GUI, etc... there are threads everywhere, even in drivers. Linux is far from that right now.
Linux has threads in the kernel, GUI, and various programs. Specifically where is it lacking support? Yes, there is room for improvement, but I can't think of any big architectural changes necessary for those improvements.
Linux does SMP too, but that doesn't mean it really takes advantage of it. You need lots of multithreading in the OS itself to really use SMP. The "SMP capable" doesn't mean "SMP efficient". On Linux you need to thread your app to take advantage of SMP. On BeOS even a non threaded app will take advantage of it because all the system calls (kernel or GUI) go to multithread code.
Indeed, you need to thread your app to make it SMP efficient as opposed to SMP capable. Unless your application spends all its time in parallelizable system calls, SMP won't help it much unless it's threaded. If you don't write your application to be threaded, you're not going to take much advantage of SMP. Good kernel SMP support takes a lot of time and resources, neither of which Be has had. It's amazing they've done as well as they have.
Linux is too command-line oriented
For me, a command line is, for simple tasks, about three times faster than a GUI, and several hundred times faster for complex tasks. However, I can run GNU/Linux without ever touching a command-line if I want. And yes, BeOS can run a command line too, so I don't see what this statement is supposed to establish.
when Unix was created graphic display was sci-fi, and when XWindows was created digital video was sci-fi too. A real multimedia OS must be built from the start to fit into today hardware (accelerated 2D, 3D, video, etc..), and have all sort of stuff to synchronize and mix audio and video. Linux is rather poor here.
"When BeOS was created, holographic displays were sci-fi." I suppose when the next newest technology comes out you'll feel the need to write another OS? I'd rather take a solid foundation and build upon it. Linux intentionally doesn't have much graphics support; they decided to put as much as possible in userland, but X that doesn't mean the OS can't synchronize and mix audio and video. Perhaps you can point at some numbers? And don't just say 'BeOS can play eight videos at the same time' because I can do that too with a lowly K6-233 and a crappy video card.
Not to say that Linux is bad, it's an excellent server OS, makes Apache and Samba rocks, but it's a very poor thing when video and audio. It was just never design to handle that. You can't just add stuff here and there to make it more "multimedia", it just makes the OS more bloated and inefficient.
Actually, adding it on modularly makes it less bloated, and the efficiency cost isn't that terrible.
You mean you have to reboot to upgrade Redhat? Bah, I think I'll stick to Debian.
You have to edit a few lines in the kernel to get it to support more than 960 MB of physical memory.
What I found interesting was that they apparently didn't make a separate swap partition for the linux box (they said 1 OS partition and 1 data partition)... hm...
But the time is signed, so it only lasts us 'till year 292277026596
: I don't think source code is any more a form of speech than a recipe
: for a cake, a set of instructions to make a bomb, or a to-do list to
: get one through a day.
And those are all protected under free speech.
By showing that the phrase 'open source' simply means 'open source' perhaps ESR will lose his battle to popularize Free software by hiding the freedom associated with it.
There's nothing ESR can do to stop their usage of the term, and in this case, Microsoft is actually correct.
Er, I'm not sure if you're joking or if you're Lewis Mettler in disguise.
You don't understand. Linux is a kernel. RMS does *not* want to rename Linux. He considers it the authors' right to call it whatever they want.
GNU is an idea, an operating system, a collection of libraries and utilities. He wants to call the combination of GNU with Linux "GNU/Linux".