Though personally I was against the war, I certainly feel Bush could have justified it with information they knew, and the fact that he lied about these WMD, and it is becoming increasing clear that he knowingly lied, is very worrysome.
Bush could easily have argued with known and proven information. Saddamm had gassed thousands of his own people, and even if he had stopped doing so for awhile, he could easily do it again. He also presented plenty of threats to his own people and his neighbors, and was probably training and harboring terrorists (just not AlQueda, who hated him as much as the US). He was taking all the money in the country and making his people starve. He was threatening the US's supply of oil, which if worded right, could be a convincing argument. He was even threatening Russia and Europe's supply of oil. He had lauched missles against Israel. There are tons of bad things that were proven about him and could be used to argue that he had to be overthrown with a military attack.
I didn't question Bush's claim that Saddamm was working on WMD. I fully expected the US to overthrow the country in a matter of weeks, and immediatly find plenty of WMD work and radioactive dumps. I am as suprised as the biggest war hawk that WMD's were not found. But, unlike me, it appears Bush had good information saying that there was a significant chance that the WMD's would not be found. Since I believe he could have easily justified the war without this, I am a rather scared of an administration that would manufacture a lie so casually to get what they want.
In case anybody is wondering how all the letters fit into 4 bits, you also have to take into account the length of the Morse code. There are in fact 2 1-bit letters, 4 2-bit ones, 8 3-bit ones, and 16 4-bit ones, for a total of 31.
Licenses are IN ADDITION TO any protection granted by copyright law. Copyright law prohibits the copying of content WITHOUT THE PERMISSION OF THE OWNER. The license grants that permission. It is entirely possible to breach the contractual provisions of a license (such as not paying royalties which are standard provsiions in most licenses) without violating copyright law.
This is where you made a mistake. The difference with the GPL is that it is impossible to break the GPL without also breaking copyright. Therefore it does not matter one iota whether the GPL is a "license" or not or even if it exists. If you want to break it, you are breaking US (and other countries) copyright laws.
The confusion is that the GPL allows you to do some things that would normally violate copyright law. Because humans are rather poor at logic, they tend to invert this and think this also means that copyright law allows you to do things the GPL does not, and thus the GPL is a contract. This is similar to the logical fallacy: "Some A's are not B" does not imply "Some B's are not A" (in fact it says nothing about whether any B's are A or not A).
"Things that violate the GPL" is a strict subset of "things that violate copyright law". Therefore there is zero need to enforce the GPL because you can rely on the far more popular, enforced, and easy-to-understand laws and regulations of copyright to completely cover it.
This is not impossible. One approach is to have government procurment require open specs of all API's, confirmed to be open and usable by a third party. Since most software and hardware companies are interested in selling to the government they would have an incentive to fulfill these requirements, without any real law over what they could do.
These specs would really have to be open and completely free. That means they could not be specified with GPL or any copyright or any license. The interesting thing about that is that if Bill Gates tries to fight such a requirement, he would find himself on the embarassing position of defending the right of somebody to GPL their work.
No, the glory days of X will be when you can assumme TrueColor and those other things exist.
We could have been there long ago if they had cleaned up their implementation, not done stupid things like "visuals" and just emulated TrueColor atop whatever hardware there was.
As it is, I still need to write colormap code in my programs. I am thinking of scrapping it, but I will need to write "select the TrueColor visual" in as well. Though a lot less objectionable, I would like this requirement to disappear as well.
Here are some things that appear to be invented by Microsoft and are important. Unfortunately this list is not going to make even Microsoft happy since everything is somewhat old, easily emulated on other systems, etc, but these IMHO are the real innovations:
1. The "taskbar", in particular the idea that a window appears in it whether or not it is "iconized" All earlier systems I have seen had the idea that a window "turned into" the icon. Thus finding an uniconized but buried window was quite difficult until they came up with this.
2. The realization that the text in the icon (taskbar) is much more imporatant than the picture. Unfortunately the HCI dweebs probably stopped them from getting rid of the icon entirely, but at least they got it very tiny.
3. The mouse wheel. Certainly the idea of the mouse being easily switched in/out of "scroll mode", or having another control on it for scrolling, has been around much longer, but they appear to have realized that limiting the idea to one dimension would allow a user-friendly solution with reliable mechanics.
4. Rasterizing graphics to the individual rgb emitters in lcd displays (what they call Cleartype and apparently only used for fonts right now by them). Yes it seems obvious now, but nobody seems to have thought of it before somebody at Microsoft did.
5. Windows 95 design that eliminated the "border line" between the window border and the contents. For instance if you drew a window containing only a gray rectangle, it merged seamlessly into the border. All earlier systems drew a divider line there. Though I know I wrote stuff like this in 1986 for the NeXT (which let you create bascally override-redirect windows), Microsoft seems to be the first commercial venture that realized you did not have to graphically seperate the edges.
There are probably other things. These are important innovations that will affect computer design long after Microsoft is irrelevant.
That's like claiming that the hundreds of thousands of creatures on Earth are unable to eat each other. Like somehow being different requires them to be different right down to their amino acids.
They are grasping at straws now. They know they are wrong.
You just said you are not prevented from doing anything, since you can rewrite the kernel. You can also lie in your module and claim it is GPL when it isn't. You can also just put those GPL_ONLY prefixes into your code and thus make it fail if somebody tries to turn on the GPL flag!
It does not seem you are prevented from doing anything.
I agree this is going to happen, but I think it will also happen at the same time as user-space drivers are added to Linux, and in fact there will be strong encouragement for all binary-only drivers to be user-space ones. This may mean that closed kernel modules are even more discouraged than they are now.
The driver does not have to be GPL you know. The company could copyright the code and say the only legal use of it is to compile it against an end-users kernel. Conversely they could also put it in the public domain.
Being able to see the source code does not imply GPL.
If this is true and taking the code home was against company policy, then that person should be in trouble anyway.
However I still suspect that this was done by somebody who had access to the code without cracking or stealing anything. They may not have had permission to see the code, and certainly no permission to upload it to the net, but I suspect there was no physical barrier (even a broken one) to them seeing the code.
Anybody stealing the code through a crack would probably have noticed and skipped the core file.They might grab everything, but could take their time to clean it up and package it before uploading it to the net, at least enough time to remove junk files. However somebody with legal access to the code would have been tempted to get it onto the net as fast as possible and would thus have sent an entire directory and missed the core file, and would not risk changing the directory or copying the files somewhere else.
The only "FUD" is some clueless readers who don't realize that you can get information out of a computer without "cracking" it. Ask them if they think their ability to read email means there is a security flaw in their computers.
Here is how the core file worked:
1. Mainsoft is certainly interested in compiling stuff on Linux, and somebody copied all this code to a Linux machine.
2. Somebody went into one of these directories and ran vim. Vim was asked to list a directory, leaving the directory list in it's memory. Vim then later crashed and left a core file containing it's entire memory in this directory.
3. Later, somebody (possibly a different person) saw this directory, said "Look at all the Microsoft source code, I think I'll post it to the net, because I am an idiot and want to risk jail time". They did not notice or care that there was a core file in the directory as well. They zipped the entire directory and posted it to the net.
3a. Alternative is that somebody legitimately backed up the direcory, copying the core file by accident. This could have gone through any number of steps and people. Eventually it reached somebody who said "hey this zip file contains Microsoft's source code, why don't I post it to the net" and it was posted. Therefore there is no proof that the owner of the machine or even a user of a Linux machine was at fault.
That quote is pretty amazing. I like his worry about a "bad release". Does he think somebody who has access to this source code can somehow make some broken version of Windows and then cause Microsoft to "release" it? What an idiot.
I don't think this is as much of a problem as people think.
Think about it: in source code you have written, do the bugs fix themselves when you look at the code? In almost all cases I have seen, what happens is the program itself crashes or does the wrong thing, and then I have to work damn hard looking at the code to try to figure out why it did so.
It is true that if you find a crack on Windows through testing, looking at the code may reveal a way to make that crack more dangerous. It might also reveal a way to easily defuse the crack. Unfortunately, unlike Linux, this code is unlikely to allow you to fix it, as apparently you cannot compile a working Windows from it, so patches cannot be tested. You could still make a suggested patch and send it to Microsoft.
"Security through obscurity" is really a term for when you write some information that you purposely keep secret. For instance a hardcoded password that you assumme won't be located by somebody looking at the compiled program, but is obvious from the code. This is obviously a problem for the released code. However you can be certain that tens of thousands of people have already been able to look at the code (Microsoft empolyeees and partners) and unless Microsoft is run by idiots they realize there are probably hundreds of them who are willing to do evil things with this information. So I really doubt that they rely on security through obscurity. This is also why I doubt there are "smoking gun" comments like "this will make Netscape slower" in the code, either. Such stuff would have been leaked already.
Duh: Mainsoft's job was to write a commercial windows-emulation for Unix. There is good reason for them to want to compile some of this code on Linux.
And this has nothing to do with a crack or security flaw. The code was leaked by somebody who had the ability to read the code anyway. Unless you think Linux's ability to retrieve information from a disk is a security flaw.
I think more likely the idiot who released this stuff took everything in a directory he found and sent it. That included a core file from when somebody ran something while in that directory and it crashed.
It is also possible that some innocent person copied the entire directory, including the core file, to another machine or tar or zip file, and the code-releasing person found that and sent it.
The core file contains enough structure to show that it was produced on a Linux machine.
None of this has anything to do with cracking or anything. It is almost certain that the guilty party had the ability to read the files without any hack.
This seems to be a popular opinion, but it is false.
You are buying into the same FUD Microsoft is spewing about the GPL.
Just looking at the code does not "taint" you. There are plenty of ex-Microsoft employees who have looked at Microsoft source code and have then contributed to non-Microsoft projects (not just OSS, but closed-source from competing companies). Really, are you claiming that a coder that has seen Microsoft's code is legally impossible to employ except at Microsoft? What if some poor sap has seen both Microsoft's code and a competitor like Suns? They can't ever work on software again anywhere?
Conversely Microsoft hires people all the time that have looked at GPL code. They don't seem worried that these people are "tainted" despite the fact that their public announcements would seem to indicate that it is impossible for such people to work there.
The person/company in trouble is the one that made the code available. Apparently this is somebody at Mainsoft, who should be punished hard. This sort of behavior is extremely damaging to IT!
In reality, looking at the code is not a problem. The person in trouble is the one who released it (and it does sound like somebody at Mainsoft did this). If looking at code was bad, you could project it on a huge wall in Times square and suddenly make hundreds of thousands of people liable to a Microsoft lawsuit! This is obviously ludricous.
There are plenty of former and even current Microsoft employees who have looked at Microsoft's code and then contributed to non-Microsoft products such as OSS and competing companies closed-source. Microsoft has not gone after them.
I remember reading that Steve Balmer and Bill Gates specifically FORBID any MS employees from reading / accessing GPL'ed code unless given express permission from somewhere high up.
That's just publicity FUD. You can be pretty certain "permission" is very easy to obtain. They are not stupid and they know quite well that the GPL does not "contaminate" but they have to make this farce announcement so the PHB's don't realize they are full of shit.
Exactly. This is what I have seen using OS/X as well. I have no idea why this is this way, and it sure does not look user-friendly to me.
I think installation on Linux is better, if it worked as the package creator intended. The problem on Linux is that the packages often don't work. Windows would be as bad if 60% of the windows installers crashed or failed with errors when you double-clicked them.
I think in the ideal system, what you get is a file that you double-click and it RUNS the program (it it is not and "installer" and not a directory containing either the program or an installer). Only if the program needs daemons or other system setup, it can then detect if it has not been installed correctly and offer to do that, or just let you run to test it. For 99% of the programs "installation" should consist of dragging that file to the correct directory so users other than yourself can see it. "uninstallation" should consist of throwing the same file in the trash, and any symbolic links or init or daemons that it "installed" should have enough smarts to delete or kill themselves when the program disappears.
Nobody (not Windows or Linux or the Mac) seem anywhere near this. Some of it seems to be complete brainwashing by the installers on Windows. We have discovered that people don't believe the installer works if it does not present them with a big scrolling box of text with an "I agree" button!
Hate to break the news to you, but both Xinerama and Windows have the same API for the monitors, pretty much a list of rectangles. Nothing prevents a program from being just as intelligent about window placement on both of them.
The main difference is that they made different decisions about the "old" one-monitor interface, used by any programs that are not aware of multiple monitors. On Windows it returns the "main" monitor, while on X it returns a rectangle around the entire set of monitors. This is likely the source of the programs that maximize or position their windows in the missing area. However conversely on Windows there are programs that refuse to work correctly if you drag them to the other monitor, for instance their overlay popups keep showing up on the main monitor.
I don't know about OS/X. The automatic update has worked great, but anything I have downloaded has arrived in a "box icon" program, and sometimes that is imbedded in a "disk image" program. I have to double-click these and then double-click the result, and it is very confusing and unclear if I can throw these things away after I have installed the program. Those virtual disks especially are confusing.
Whether it is safe to throw away the RPM's is also unclear to me. For some reason Windows installers do seem to make it clear that you can throw away the install program after using it.
Incidentally all the Linux desktops I have ever seen let you double-click RPM's and they *try* to install them. The problem is not that you can't do this, but that all too often it does not work, and you are forced to go to the shell to try again with the "--force" switch or whatever. I suppose you could make an argument that if a Windows installer does not work you are completely hosed, while it is physically possible to fix an RPM, but in reality Windows installers tend to always work (only counter example I have seen was an ATI driver for an old OpenGL card that crashed and the driver did not appear until I rebooted the machine).
I am still baffled why Windows has so brainwashed people that they think they need to "install" anything. Really I should be able to grab the file from the web page, drop it on the desktop, double-click it, and RUN the program, not "install" it. If I want it on the start menu I should then drag it there. And to "uninstall" I should be able to drag it and drop it in the trash.
Programs that need to mess with.rc files or whatever should run a demo mode and pop up a dialog that says "if this seems to be doing what you want, click here to make changes to your system so it really works..."
Getting all the library and system calls to work would allow Mac programs to run under this on an x86 by recompiling them and linking with the emulation libraries that they develop, just like Wine is supposed to allow you to recompile Windows programs to run them under Linux on other platforms.
At least that is the idea. I agree that in reality it is not working out that way, it appears a lot of Wine is "get this DLL to run". Even a company making a Windows program and interested in porting it to Linux and in recompiling it cannot use Wine to do this, as they still rely on some closed-source DLL that is not emulated and is provided by a company like Microsoft who definatly will not recompile it.
Don't forget that M$ led a drive to make "high definition" television 640x480...
First of all, existing NTSC analog is considered by the D1 encoding standard to be 486 lines tall and 720 pixels wide. The 486 tall was taken as being the useful portion of the actual 525 scan lines in the signal (the rest are synchronization and retrace intervals). The 720 horizontal was decided on as being approximately the resolution of the analog signal.
Because it is a multiple of 16 and a multiple of the 24 lines of text that were often displayed the size 480 was also chosen for lots of early computer displays, and was used on the IBM PC monochrome display. The horizontal resolution of 640 was chosen so that a 4/3 aspect ratio could be displayed with square pixels. This is pretty common now for the lowest resolution a PC display card can be set to.
From what I have heard, Microsoft and almost all computer software and hardware manufacturers pushed for much better HDTV standards, such as 1024p. It was broadcasters that insisted on interlaced and lower-resolution standards. Even if you assumme evil intentions on Microsoft's part, they certainly would want as high a resolution as possible so that people are forced to upgrade and thus buy new software in order to display HDTV.
Slashdot Mirror
Though personally I was against the war, I certainly feel Bush could have justified it with information they knew, and the fact that he lied about these WMD, and it is becoming increasing clear that he knowingly lied, is very worrysome.
Bush could easily have argued with known and proven information. Saddamm had gassed thousands of his own people, and even if he had stopped doing so for awhile, he could easily do it again. He also presented plenty of threats to his own people and his neighbors, and was probably training and harboring terrorists (just not AlQueda, who hated him as much as the US). He was taking all the money in the country and making his people starve. He was threatening the US's supply of oil, which if worded right, could be a convincing argument. He was even threatening Russia and Europe's supply of oil. He had lauched missles against Israel. There are tons of bad things that were proven about him and could be used to argue that he had to be overthrown with a military attack.
I didn't question Bush's claim that Saddamm was working on WMD. I fully expected the US to overthrow the country in a matter of weeks, and immediatly find plenty of WMD work and radioactive dumps. I am as suprised as the biggest war hawk that WMD's were not found. But, unlike me, it appears Bush had good information saying that there was a significant chance that the WMD's would not be found. Since I believe he could have easily justified the war without this, I am a rather scared of an administration that would manufacture a lie so casually to get what they want.
In case anybody is wondering how all the letters fit into 4 bits, you also have to take into account the length of the Morse code. There are in fact 2 1-bit letters, 4 2-bit ones, 8 3-bit ones, and 16 4-bit ones, for a total of 31.
Licenses are IN ADDITION TO any protection granted by copyright law. Copyright law prohibits the copying of content WITHOUT THE PERMISSION OF THE OWNER. The license grants that permission. It is entirely possible to breach the contractual provisions of a license (such as not paying royalties which are standard provsiions in most licenses) without violating copyright law.
This is where you made a mistake. The difference with the GPL is that it is impossible to break the GPL without also breaking copyright. Therefore it does not matter one iota whether the GPL is a "license" or not or even if it exists. If you want to break it, you are breaking US (and other countries) copyright laws.
The confusion is that the GPL allows you to do some things that would normally violate copyright law. Because humans are rather poor at logic, they tend to invert this and think this also means that copyright law allows you to do things the GPL does not, and thus the GPL is a contract. This is similar to the logical fallacy: "Some A's are not B" does not imply "Some B's are not A" (in fact it says nothing about whether any B's are A or not A).
"Things that violate the GPL" is a strict subset of "things that violate copyright law". Therefore there is zero need to enforce the GPL because you can rely on the far more popular, enforced, and easy-to-understand laws and regulations of copyright to completely cover it.
This is not impossible. One approach is to have government procurment require open specs of all API's, confirmed to be open and usable by a third party. Since most software and hardware companies are interested in selling to the government they would have an incentive to fulfill these requirements, without any real law over what they could do.
These specs would really have to be open and completely free. That means they could not be specified with GPL or any copyright or any license. The interesting thing about that is that if Bill Gates tries to fight such a requirement, he would find himself on the embarassing position of defending the right of somebody to GPL their work.
He made this promise after MyDoom was out there, and in fact in the very same speech in which he talked about MyDoom.
I expect the excuse they will come up with is to say that IBM is a linux user and they are already suing them.
No, the glory days of X will be when you can assumme TrueColor and those other things exist.
We could have been there long ago if they had cleaned up their implementation, not done stupid things like "visuals" and just emulated TrueColor atop whatever hardware there was.
As it is, I still need to write colormap code in my programs. I am thinking of scrapping it, but I will need to write "select the TrueColor visual" in as well. Though a lot less objectionable, I would like this requirement to disappear as well.
Here are some things that appear to be invented by Microsoft and are important. Unfortunately this list is not going to make even Microsoft happy since everything is somewhat old, easily emulated on other systems, etc, but these IMHO are the real innovations:
1. The "taskbar", in particular the idea that a window appears in it whether or not it is "iconized" All earlier systems I have seen had the idea that a window "turned into" the icon. Thus finding an uniconized but buried window was quite difficult until they came up with this.
2. The realization that the text in the icon (taskbar) is much more imporatant than the picture. Unfortunately the HCI dweebs probably stopped them from getting rid of the icon entirely, but at least they got it very tiny.
3. The mouse wheel. Certainly the idea of the mouse being easily switched in/out of "scroll mode", or having another control on it for scrolling, has been around much longer, but they appear to have realized that limiting the idea to one dimension would allow a user-friendly solution with reliable mechanics.
4. Rasterizing graphics to the individual rgb emitters in lcd displays (what they call Cleartype and apparently only used for fonts right now by them). Yes it seems obvious now, but nobody seems to have thought of it before somebody at Microsoft did.
5. Windows 95 design that eliminated the "border line" between the window border and the contents. For instance if you drew a window containing only a gray rectangle, it merged seamlessly into the border. All earlier systems drew a divider line there. Though I know I wrote stuff like this in 1986 for the NeXT (which let you create bascally override-redirect windows), Microsoft seems to be the first commercial venture that realized you did not have to graphically seperate the edges.
There are probably other things. These are important innovations that will affect computer design long after Microsoft is irrelevant.
That's like claiming that the hundreds of thousands of creatures on Earth are unable to eat each other. Like somehow being different requires them to be different right down to their amino acids.
They are grasping at straws now. They know they are wrong.
You just said you are not prevented from doing anything, since you can rewrite the kernel. You can also lie in your module and claim it is GPL when it isn't. You can also just put those GPL_ONLY prefixes into your code and thus make it fail if somebody tries to turn on the GPL flag!
It does not seem you are prevented from doing anything.
I agree this is going to happen, but I think it will also happen at the same time as user-space drivers are added to Linux, and in fact there will be strong encouragement for all binary-only drivers to be user-space ones. This may mean that closed kernel modules are even more discouraged than they are now.
The driver does not have to be GPL you know. The company could copyright the code and say the only legal use of it is to compile it against an end-users kernel. Conversely they could also put it in the public domain.
Being able to see the source code does not imply GPL.
If this is true and taking the code home was against company policy, then that person should be in trouble anyway.
However I still suspect that this was done by somebody who had access to the code without cracking or stealing anything. They may not have had permission to see the code, and certainly no permission to upload it to the net, but I suspect there was no physical barrier (even a broken one) to them seeing the code.
Anybody stealing the code through a crack would probably have noticed and skipped the core file.They might grab everything, but could take their time to clean it up and package it before uploading it to the net, at least enough time to remove junk files. However somebody with legal access to the code would have been tempted to get it onto the net as fast as possible and would thus have sent an entire directory and missed the core file, and would not risk changing the directory or copying the files somewhere else.
The only "FUD" is some clueless readers who don't realize that you can get information out of a computer without "cracking" it. Ask them if they think their ability to read email means there is a security flaw in their computers.
Here is how the core file worked:
1. Mainsoft is certainly interested in compiling stuff on Linux, and somebody copied all this code to a Linux machine.
2. Somebody went into one of these directories and ran vim. Vim was asked to list a directory, leaving the directory list in it's memory. Vim then later crashed and left a core file containing it's entire memory in this directory.
3. Later, somebody (possibly a different person) saw this directory, said "Look at all the Microsoft source code, I think I'll post it to the net, because I am an idiot and want to risk jail time". They did not notice or care that there was a core file in the directory as well. They zipped the entire directory and posted it to the net.
3a. Alternative is that somebody legitimately backed up the direcory, copying the core file by accident. This could have gone through any number of steps and people. Eventually it reached somebody who said "hey this zip file contains Microsoft's source code, why don't I post it to the net" and it was posted. Therefore there is no proof that the owner of the machine or even a user of a Linux machine was at fault.
That quote is pretty amazing. I like his worry about a "bad release". Does he think somebody who has access to this source code can somehow make some broken version of Windows and then cause Microsoft to "release" it? What an idiot.
I don't think this is as much of a problem as people think.
Think about it: in source code you have written, do the bugs fix themselves when you look at the code? In almost all cases I have seen, what happens is the program itself crashes or does the wrong thing, and then I have to work damn hard looking at the code to try to figure out why it did so.
It is true that if you find a crack on Windows through testing, looking at the code may reveal a way to make that crack more dangerous. It might also reveal a way to easily defuse the crack. Unfortunately, unlike Linux, this code is unlikely to allow you to fix it, as apparently you cannot compile a working Windows from it, so patches cannot be tested. You could still make a suggested patch and send it to Microsoft.
"Security through obscurity" is really a term for when you write some information that you purposely keep secret. For instance a hardcoded password that you assumme won't be located by somebody looking at the compiled program, but is obvious from the code. This is obviously a problem for the released code. However you can be certain that tens of thousands of people have already been able to look at the code (Microsoft empolyeees and partners) and unless Microsoft is run by idiots they realize there are probably hundreds of them who are willing to do evil things with this information. So I really doubt that they rely on security through obscurity. This is also why I doubt there are "smoking gun" comments like "this will make Netscape slower" in the code, either. Such stuff would have been leaked already.
Duh: Mainsoft's job was to write a commercial windows-emulation for Unix. There is good reason for them to want to compile some of this code on Linux.
And this has nothing to do with a crack or security flaw. The code was leaked by somebody who had the ability to read the code anyway. Unless you think Linux's ability to retrieve information from a disk is a security flaw.
I think more likely the idiot who released this stuff took everything in a directory he found and sent it. That included a core file from when somebody ran something while in that directory and it crashed.
It is also possible that some innocent person copied the entire directory, including the core file, to another machine or tar or zip file, and the code-releasing person found that and sent it.
The core file contains enough structure to show that it was produced on a Linux machine.
None of this has anything to do with cracking or anything. It is almost certain that the guilty party had the ability to read the files without any hack.
This seems to be a popular opinion, but it is false.
You are buying into the same FUD Microsoft is spewing about the GPL.
Just looking at the code does not "taint" you. There are plenty of ex-Microsoft employees who have looked at Microsoft source code and have then contributed to non-Microsoft projects (not just OSS, but closed-source from competing companies). Really, are you claiming that a coder that has seen Microsoft's code is legally impossible to employ except at Microsoft? What if some poor sap has seen both Microsoft's code and a competitor like Suns? They can't ever work on software again anywhere?
Conversely Microsoft hires people all the time that have looked at GPL code. They don't seem worried that these people are "tainted" despite the fact that their public announcements would seem to indicate that it is impossible for such people to work there.
The person/company in trouble is the one that made the code available. Apparently this is somebody at Mainsoft, who should be punished hard. This sort of behavior is extremely damaging to IT!
Copying the code is definately illegal.
In reality, looking at the code is not a problem. The person in trouble is the one who released it (and it does sound like somebody at Mainsoft did this). If looking at code was bad, you could project it on a huge wall in Times square and suddenly make hundreds of thousands of people liable to a Microsoft lawsuit! This is obviously ludricous.
There are plenty of former and even current Microsoft employees who have looked at Microsoft's code and then contributed to non-Microsoft products such as OSS and competing companies closed-source. Microsoft has not gone after them.
I remember reading that Steve Balmer and Bill Gates specifically FORBID any MS employees from reading / accessing GPL'ed code unless given express permission from somewhere high up.
That's just publicity FUD. You can be pretty certain "permission" is very easy to obtain. They are not stupid and they know quite well that the GPL does not "contaminate" but they have to make this farce announcement so the PHB's don't realize they are full of shit.
Exactly. This is what I have seen using OS/X as well. I have no idea why this is this way, and it sure does not look user-friendly to me.
I think installation on Linux is better, if it worked as the package creator intended. The problem on Linux is that the packages often don't work. Windows would be as bad if 60% of the windows installers crashed or failed with errors when you double-clicked them.
I think in the ideal system, what you get is a file that you double-click and it RUNS the program (it it is not and "installer" and not a directory containing either the program or an installer). Only if the program needs daemons or other system setup, it can then detect if it has not been installed correctly and offer to do that, or just let you run to test it. For 99% of the programs "installation" should consist of dragging that file to the correct directory so users other than yourself can see it. "uninstallation" should consist of throwing the same file in the trash, and any symbolic links or init or daemons that it "installed" should have enough smarts to delete or kill themselves when the program disappears.
Nobody (not Windows or Linux or the Mac) seem anywhere near this. Some of it seems to be complete brainwashing by the installers on Windows. We have discovered that people don't believe the installer works if it does not present them with a big scrolling box of text with an "I agree" button!
Hate to break the news to you, but both Xinerama and Windows have the same API for the monitors, pretty much a list of rectangles. Nothing prevents a program from being just as intelligent about window placement on both of them.
The main difference is that they made different decisions about the "old" one-monitor interface, used by any programs that are not aware of multiple monitors. On Windows it returns the "main" monitor, while on X it returns a rectangle around the entire set of monitors. This is likely the source of the programs that maximize or position their windows in the missing area. However conversely on Windows there are programs that refuse to work correctly if you drag them to the other monitor, for instance their overlay popups keep showing up on the main monitor.
I don't know about OS/X. The automatic update has worked great, but anything I have downloaded has arrived in a "box icon" program, and sometimes that is imbedded in a "disk image" program. I have to double-click these and then double-click the result, and it is very confusing and unclear if I can throw these things away after I have installed the program. Those virtual disks especially are confusing.
.rc files or whatever should run a demo mode and pop up a dialog that says "if this seems to be doing what you want, click here to make changes to your system so it really works..."
Whether it is safe to throw away the RPM's is also unclear to me. For some reason Windows installers do seem to make it clear that you can throw away the install program after using it.
Incidentally all the Linux desktops I have ever seen let you double-click RPM's and they *try* to install them. The problem is not that you can't do this, but that all too often it does not work, and you are forced to go to the shell to try again with the "--force" switch or whatever. I suppose you could make an argument that if a Windows installer does not work you are completely hosed, while it is physically possible to fix an RPM, but in reality Windows installers tend to always work (only counter example I have seen was an ATI driver for an old OpenGL card that crashed and the driver did not appear until I rebooted the machine).
I am still baffled why Windows has so brainwashed people that they think they need to "install" anything. Really I should be able to grab the file from the web page, drop it on the desktop, double-click it, and RUN the program, not "install" it. If I want it on the start menu I should then drag it there. And to "uninstall" I should be able to drag it and drop it in the trash.
Programs that need to mess with
Getting all the library and system calls to work would allow Mac programs to run under this on an x86 by recompiling them and linking with the emulation libraries that they develop, just like Wine is supposed to allow you to recompile Windows programs to run them under Linux on other platforms.
At least that is the idea. I agree that in reality it is not working out that way, it appears a lot of Wine is "get this DLL to run". Even a company making a Windows program and interested in porting it to Linux and in recompiling it cannot use Wine to do this, as they still rely on some closed-source DLL that is not emulated and is provided by a company like Microsoft who definatly will not recompile it.
Don't forget that M$ led a drive to make "high definition" television 640x480...
First of all, existing NTSC analog is considered by the D1 encoding standard to be 486 lines tall and 720 pixels wide. The 486 tall was taken as being the useful portion of the actual 525 scan lines in the signal (the rest are synchronization and retrace intervals). The 720 horizontal was decided on as being approximately the resolution of the analog signal.
Because it is a multiple of 16 and a multiple of the 24 lines of text that were often displayed the size 480 was also chosen for lots of early computer displays, and was used on the IBM PC monochrome display. The horizontal resolution of 640 was chosen so that a 4/3 aspect ratio could be displayed with square pixels. This is pretty common now for the lowest resolution a PC display card can be set to.
From what I have heard, Microsoft and almost all computer software and hardware manufacturers pushed for much better HDTV standards, such as 1024p. It was broadcasters that insisted on interlaced and lower-resolution standards. Even if you assumme evil intentions on Microsoft's part, they certainly would want as high a resolution as possible so that people are forced to upgrade and thus buy new software in order to display HDTV.