And to make sure the RIAA doesn't get your money, you can return the CD after!
I have been considering that idea before. Is that even illegal? Where I live making a copy for personal use is legal as long as the original is legal. And then we can return the original afterwards because it is corrupt. And since all the other discs they can find will also be corrupt, they will have no choice but paying the money back.
where do you suppose the MP3's will be coming from if all the CDs have copy protection?
If the "CD" can be played it can be copied. It is as simple as that, there is no such thing as copy protection. What they usually do is make errors on the disc and hope for devices used for copying to be more sensitive to errors than players.
I bet they are not all on that list. The one I got for christmas says "COPY CONTROL" on the cover. So I assume it is YACCD (yet another corrupt CD), but it is not on the list. But then again, I didn't have time to test it yet.
Do they have to fall for it? Sign the NDA, get the specs, learn the relevant parts by heart, go on vacation in a country where the NDA does not apply, and tell somebody what you know. They wouldn't know before it is too late, that you were going to reveal anything. And you probably wouldn't have done anything illegal. And finally they might even have a hard time finding out who did it.
Microsoft can sign NDAs. Open source developers can't.
Maybe we can find a country where a developer can legally ignore the NDA he signed.
Re:This is the reason for Windows's advantage
on
DirectX 9 Finally Out
·
· Score: 2, Insightful
It is hardly any surprise that Microsoft developers with access to specs support the hardware earlier than opensource developers without access to the specs.
has Linux actually improved over the last year in any real way?
Yes, great things has happened in the development of 2.5.
I have a feeling better things will come in 2003.
I guess your feeling will prove correct. To be a litle more specific I expect the 2.5 development to finish and we will see a new stable kernel with the improvements we are looking forward for.
understand the source code enough to implement the 0xCD interrupt.
The problem is not understanding the source but rather understanding what the 0xCD interrupt is supposed to do. In fact I have never heard about that interrupt, so maybe the bug is not really that it is not implemented, but rather that it is being called at all.
What about IBM's queue reordering with their new 180gxp drives?
That is indeed an interesting question. Does this converter actully support that feature? And can an IDE drive with such a converter perform as good as a SCSI drive, assuming both drive and converter is implemented correctlly?
I'm sure it is possible to create an IDE controller that can match the performance of this converter connected to the best of all SCSI controllers. So the only reasons for prefering this converter over a good IDE controller are:
Redhat and suse both ship 9 month old hacked to deat kernels while debian has 2.4.19.
What is the difference? Surely RedHat has included a lot of the same patches that has went into 2.4.19. And IIRC RedHat 7.3 was released before 2.4.19, so they couldn't possibly have used anything newer than the 2.4.18 they picked from the stable series.
Drivers don't have to resolve every symbol possible do they?
No they don't have to. But there are other ways the ABI can break. A symbol being exported or not is the least significant change of the ABI, and that change can easilly be reverted. What is much more important is the layout of structs used in the kernel. A struct with same name in two different kernel builds does not necesarilly have the same fields at the same offsets or even the same size.
normally, the GPL doesn't allow for non-source availible binaries to be linked into GPL code
It is a very grey area. The GPL does not forbid linking GPL code with closed source code for your own personal use. But you are not allowed to distribute such code.
Now since the modules as they are distributed is not yet linked they might be legal. The linking is done at load time, at which time it might be considered personal use.
What commercial application would you most like to see migrated to run natively on Linux"? I guess for these kernel developers the answer is "NONE of them".
There are major differences between kernel modules and applications. I don't use any closed source kernel modules, and I certainly hope I never have to. But I happily use a few closed source userspace programs.
If you take a closer look on the kernel/module ABI and the kernel/user mode ABI, you will see that the first one is changing basically on every build of the kernel while the second is almost unchanged all the way from 2.0 to 2.6.
Closed source modules never was a good idea and always causes trouble, while userspace applications can be closed source and work on a wide range of kernels because the system call ABI is very stable.
I don't know exactly what a Winmodem is. But from what I have heard it is a lousy poorly designed hardware component that is best used for slowing down the system. The purpose of the design was to cut a few cents off the price for the component while at the same time making the hardware a lot worse. Work that was best done in hardware now has to be done in software because the component doing it was removed. But many people don't know about all this, they think all modems are equal. They think the clockfrequency of the CPU is the most important factor on system speed, and they surely don't know how many clockcycles will be wasted due to the Winmodem. A Winmodem is a piece of hardware so badly designed that you cannot write a good driver for it.
Why does the kernel module interface change so often though? I mean, surely there are only so many ways to write a mouse driver
The mouse driver is one of the simplest drivers, and most of the stuff is done in usermode anyway. But a changing interface doesn't mean that each and every piece of the interface changes, it just means that the overall interface has changed in some way.
It doesn't take much to make a change in the binary interface, a small change of a.h file will do. In fact even changing just a kernel configuration option could change the binary interface. The API on source level is however more stable. It doesn't change that often, and it doesn't change in unpredictable ways by modifying something completely unrelated to what you are writing. This API has been kept almost unchanged all the way through the 2.4 series.
You left off one major con:
Binary-only modules must be updated by the vendor when the kernel interface changes.
That is not just a major con, it is probably the most important of all cons.
you compile a small wrapper function with a provided library file to generate the module. That way you can generate one specific to your kernel version.
Hasn't something similar been done for gfx drivers as well?
I don't believe it. Assume a driver has been written and released in binary form. How could the driver become any worse by releasing the source? It couldn't, but more people might realize how badly it was written. And nVidia's is assumed responsible for a lot of the kernel crashes. If anybody reports and error on a system that has been using nVidia drivers, the kernel developers will tell the person to reproduce it without the nVidia driver or go bother nVidia.
Why is there still no standard model for adding and removing apps?
There are a few different of which two or three are very widely used. Of course having more than one is a litle unfortunate, but there are actually multiple of those which are quite good. Even with multiple different aproaches, it is still better than what you see on other systems. On other systems you basically have one installer program for each application which is a lot more than an installer program for each distribution.
different desktops, ALL of which are mutually incompatible?
They are all built on top of X11. And they are not all incompatible, I frequently mix applications from different environments. Only a few applications are dependend on a particular environment.
The lack of standards in Linux is even worse than the closed-ended standards on other OSes (coughWindowscough) because it makes almost any attempt to converge standards nearly impossible.
I completely disagree with that. With Windows you don't get any kind of standard. The MS way of compatibility is not achieved through standards, but rather by having only one implementation. And they can move it wherever they want. Yesterdays version of Windows is not necesarilly compatible with tomorrows. And if MS dislikes you, your applications will be the ones to suffer the most. And finally tomorrows version of Windows will be incompatible with todays competitors.
KDE, Gnome, I don't care. Pick one and use it.
Sure, a lot of people do that. They just don't pick the same. And who do you think is in a position to deside which of the two people are allowed to use? We are talking about freedom here, people cannot come and tell me I must use gnome or I must use KDE. I often alternate between gnome and KDE, whenever I upgrade I use the one giving the best performance and stability in that particular distribution. Or I even use twm if I get too tired with the whole thing, in fact KDE and Gnome are becoming too much bloatware for my taste. I liked them better in the old days, if they would just have worked on the stability rather than the bloatness it would have been so much better today.
Face it -- GUIs make your life easier
Sure, I sometimes want to use a GUI. But I don't need the entire desktop environment concept known from Gnome and KDE. Give me a Window Manager and nothing more. All I need is a nice way to manage the Windows on my screen, and of course a way to open new xterms.
AFAIK it is still the best OS available for a 286. But then again, who still uses 286? (... Wait a minute, my terminal emulator runs on a 286 with DOS.)
just to flash the BIOS... but novice users do not want to open the computer for fear of breaking something.
I don't consider that a problem. Even as an experienced user I would fear breaking something when flashing the BIOS. I'd expect larger risk of breaking something by flashing the BIOS than I would from opening the computer and moving a jumper.
But if you really want the jumper easily accessible, the pins could be connected to a button on the case. I don't think many new motherboards uses the turbo button for anything.
The black hats have trojaned, and will continue to trojan, a machine's flash BIOS.
It is IMHO a bug on the MOBO if the BIOS by default can be flashed, it should be required at least to move a jumper to flash it. Another interesting approach would be to have a MOBO with two flash BIOSes, only one is writable, and the other is the one being used during boot. The roles of the two can be switched by a jumper.
Then your compromised system might apear to be clean. I have actually seen a system where that has happened. But the intruder forgot to trojan the rpm executable, "rpm -Va" revealed everything. But had the intruder trojaned the rpm executable too, that wouldn't have worked. The only secure way to use the verification tool is to boot from a readonly media and run the tool from there.
Slashdot Mirror
And to make sure the RIAA doesn't get your money, you can return the CD after!
I have been considering that idea before. Is that even illegal? Where I live making a copy for personal use is legal as long as the original is legal. And then we can return the original afterwards because it is corrupt. And since all the other discs they can find will also be corrupt, they will have no choice but paying the money back.
where do you suppose the MP3's will be coming from if all the CDs have copy protection?
If the "CD" can be played it can be copied. It is as simple as that, there is no such thing as copy protection. What they usually do is make errors on the disc and hope for devices used for copying to be more sensitive to errors than players.
They're all listed
I bet they are not all on that list. The one I got for christmas says "COPY CONTROL" on the cover. So I assume it is YACCD (yet another corrupt CD), but it is not on the list. But then again, I didn't have time to test it yet.
Exactly how many Danes post to slashdot?
I do, and I know I'm not the only one. But I cannot tell you exactly how many we are.
I have a feeling they wouldn't fall for it.
Do they have to fall for it? Sign the NDA, get the specs, learn the relevant parts by heart, go on vacation in a country where the NDA does not apply, and tell somebody what you know. They wouldn't know before it is too late, that you were going to reveal anything. And you probably wouldn't have done anything illegal. And finally they might even have a hard time finding out who did it.
Microsoft can sign NDAs. Open source developers can't.
Maybe we can find a country where a developer can legally ignore the NDA he signed.
It is hardly any surprise that Microsoft developers with access to specs support the hardware earlier than opensource developers without access to the specs.
has Linux actually improved over the last year in any real way?
Yes, great things has happened in the development of 2.5.
I have a feeling better things will come in 2003.
I guess your feeling will prove correct. To be a litle more specific I expect the 2.5 development to finish and we will see a new stable kernel with the improvements we are looking forward for.
understand the source code enough to implement the 0xCD interrupt.
The problem is not understanding the source but rather understanding what the 0xCD interrupt is supposed to do. In fact I have never heard about that interrupt, so maybe the bug is not really that it is not implemented, but rather that it is being called at all.
That is indeed an interesting question. Does this converter actully support that feature? And can an IDE drive with such a converter perform as good as a SCSI drive, assuming both drive and converter is implemented correctlly?
I'm sure it is possible to create an IDE controller that can match the performance of this converter connected to the best of all SCSI controllers. So the only reasons for prefering this converter over a good IDE controller are:
Redhat and suse both ship 9 month old hacked to deat kernels while debian has 2.4.19.
What is the difference? Surely RedHat has included a lot of the same patches that has went into 2.4.19. And IIRC RedHat 7.3 was released before 2.4.19, so they couldn't possibly have used anything newer than the 2.4.18 they picked from the stable series.
Drivers don't have to resolve every symbol possible do they?
No they don't have to. But there are other ways the ABI can break. A symbol being exported or not is the least significant change of the ABI, and that change can easilly be reverted. What is much more important is the layout of structs used in the kernel. A struct with same name in two different kernel builds does not necesarilly have the same fields at the same offsets or even the same size.
normally, the GPL doesn't allow for non-source availible binaries to be linked into GPL code
It is a very grey area. The GPL does not forbid linking GPL code with closed source code for your own personal use. But you are not allowed to distribute such code.
Now since the modules as they are distributed is not yet linked they might be legal. The linking is done at load time, at which time it might be considered personal use.
What commercial application would you most like to see migrated to run natively on Linux"? I guess for these kernel developers the answer is "NONE of them".
There are major differences between kernel modules and applications. I don't use any closed source kernel modules, and I certainly hope I never have to. But I happily use a few closed source userspace programs.
If you take a closer look on the kernel/module ABI and the kernel/user mode ABI, you will see that the first one is changing basically on every build of the kernel while the second is almost unchanged all the way from 2.0 to 2.6.
Closed source modules never was a good idea and always causes trouble, while userspace applications can be closed source and work on a wide range of kernels because the system call ABI is very stable.
I don't know exactly what a Winmodem is. But from what I have heard it is a lousy poorly designed hardware component that is best used for slowing down the system. The purpose of the design was to cut a few cents off the price for the component while at the same time making the hardware a lot worse. Work that was best done in hardware now has to be done in software because the component doing it was removed. But many people don't know about all this, they think all modems are equal. They think the clockfrequency of the CPU is the most important factor on system speed, and they surely don't know how many clockcycles will be wasted due to the Winmodem. A Winmodem is a piece of hardware so badly designed that you cannot write a good driver for it.
Why does the kernel module interface change so often though? I mean, surely there are only so many ways to write a mouse driver
.h file will do. In fact even changing just a kernel configuration option could change the binary interface. The API on source level is however more stable. It doesn't change that often, and it doesn't change in unpredictable ways by modifying something completely unrelated to what you are writing. This API has been kept almost unchanged all the way through the 2.4 series.
The mouse driver is one of the simplest drivers, and most of the stuff is done in usermode anyway. But a changing interface doesn't mean that each and every piece of the interface changes, it just means that the overall interface has changed in some way.
It doesn't take much to make a change in the binary interface, a small change of a
You left off one major con:
Binary-only modules must be updated by the vendor when the kernel interface changes.
That is not just a major con, it is probably the most important of all cons.
you compile a small wrapper function with a provided library file to generate the module. That way you can generate one specific to your kernel version.
Hasn't something similar been done for gfx drivers as well?
Better drivers (usually)
I don't believe it. Assume a driver has been written and released in binary form. How could the driver become any worse by releasing the source? It couldn't, but more people might realize how badly it was written. And nVidia's is assumed responsible for a lot of the kernel crashes. If anybody reports and error on a system that has been using nVidia drivers, the kernel developers will tell the person to reproduce it without the nVidia driver or go bother nVidia.
Why is there still no standard model for adding and removing apps?
There are a few different of which two or three are very widely used. Of course having more than one is a litle unfortunate, but there are actually multiple of those which are quite good. Even with multiple different aproaches, it is still better than what you see on other systems. On other systems you basically have one installer program for each application which is a lot more than an installer program for each distribution.
different desktops, ALL of which are mutually incompatible?
They are all built on top of X11. And they are not all incompatible, I frequently mix applications from different environments. Only a few applications are dependend on a particular environment.
The lack of standards in Linux is even worse than the closed-ended standards on other OSes (coughWindowscough) because it makes almost any attempt to converge standards nearly impossible.
I completely disagree with that. With Windows you don't get any kind of standard. The MS way of compatibility is not achieved through standards, but rather by having only one implementation. And they can move it wherever they want. Yesterdays version of Windows is not necesarilly compatible with tomorrows. And if MS dislikes you, your applications will be the ones to suffer the most. And finally tomorrows version of Windows will be incompatible with todays competitors.
KDE, Gnome, I don't care. Pick one and use it.
Sure, a lot of people do that. They just don't pick the same. And who do you think is in a position to deside which of the two people are allowed to use? We are talking about freedom here, people cannot come and tell me I must use gnome or I must use KDE. I often alternate between gnome and KDE, whenever I upgrade I use the one giving the best performance and stability in that particular distribution. Or I even use twm if I get too tired with the whole thing, in fact KDE and Gnome are becoming too much bloatware for my taste. I liked them better in the old days, if they would just have worked on the stability rather than the bloatness it would have been so much better today.
Face it -- GUIs make your life easier
Sure, I sometimes want to use a GUI. But I don't need the entire desktop environment concept known from Gnome and KDE. Give me a Window Manager and nothing more. All I need is a nice way to manage the Windows on my screen, and of course a way to open new xterms.
Linus contributes the kernel, which is a great thing, but still is only a part of any distribution.
Sad seing such an Insightful comment only scoring 0. Unfortunately I just used my last moderation point.
but who can guarantee me that the binaries don't have those backdoors and trojans.
Take that one step further and read Ken Thompson's masterpiece Reflections on Trusting Trust
Who hell still uses this dinosaur anyhow?
AFAIK it is still the best OS available for a 286. But then again, who still uses 286? (... Wait a minute, my terminal emulator runs on a 286 with DOS.)
just to flash the BIOS... but novice users do not want to open the computer for fear of breaking something.
I don't consider that a problem. Even as an experienced user I would fear breaking something when flashing the BIOS. I'd expect larger risk of breaking something by flashing the BIOS than I would from opening the computer and moving a jumper.
But if you really want the jumper easily accessible, the pins could be connected to a button on the case. I don't think many new motherboards uses the turbo button for anything.
The black hats have trojaned, and will continue to trojan, a machine's flash BIOS.
It is IMHO a bug on the MOBO if the BIOS by default can be flashed, it should be required at least to move a jumper to flash it. Another interesting approach would be to have a MOBO with two flash BIOSes, only one is writable, and the other is the one being used during boot. The roles of the two can be switched by a jumper.
when they trojan your MD5 checksummer?
Then your compromised system might apear to be clean. I have actually seen a system where that has happened. But the intruder forgot to trojan the rpm executable, "rpm -Va" revealed everything. But had the intruder trojaned the rpm executable too, that wouldn't have worked. The only secure way to use the verification tool is to boot from a readonly media and run the tool from there.