This is in OpenGL shading language, which is probably in a string constant in your C code, so there are no conflicts.Ie the C code is something like gl_set_shader(gl_compile_shader("the source code goes here"));
If you want to cut & paste the code so the same algorithim can be used by both your C program and shaders you can use macros. Ie "#define gl_vec2 vec2" can be inserted into the shader language fragment and then you can use gl_vec2 the way you want.
There is a world of difference between encryption and DRM. Encryption relies on two parties both of who are interested in keeping a secret. DRM attempts to make a party that is NOT interested in keeping a secret to do so. This violates the laws of physics and causality and is in the end impossible. Attempting to make the impossible possible will just make things incredibly inconvienent. The fact that it also makes competition with Microsoft impossible also makes everybody really pretty upset.
Good one! The HDR code is available for commercial use if you negotiate a license with Digital Domain. Sort of like Qt. Also the algorithim is simple enough that you can probably recreate it without looking at the code.
The direct answer is that yes you will be able to do anything with the code after the copyright expires. GPL is a license to violate the copyright in certain restricted cases, so if the copyright is expired the GPL is meaningless.
I agree with all the joke responses that this is probably never going to happen in the USA.
Wrong. Nearly every useful library on Linux is LGPL already (the only counterexample anybody seems to be able to come up with is readline, I would like to hear what the others are).
The reason commercial software is statically-linked is that they are afraid of binary incompatabilities. Those static-linked libraries are the same ones.
I think the answer is that although the GPL could in theory lead to perverse examples, this never happens in the real world.
It is obvious that the DNS server was written to allow closed-source software to call it. Otherwise it would be quite useless, people would not use it, or ignore the restriction (because it is impossible to enforce technically). I think all the really perverse examples that can be come up with it can be shown quite obviously that the writer intended non-GNU software to call it.
As for "libraries", it seems that all real-world examples are LGPL or similar. The only counter-example I have ever heard of is readline, can anybody (even RMS) name another? By this I mean a library that could in any concievable way be used by more than one program with significantly different functions. Libraries designed to work around the Windows DLL limitations so that "plugins" can load do not count.
In the real world (as opposed to RMS's fantasy wourld) the LGPL (or an even looser version that explicitly allows static linking, like I use) very well protects your clever algorithim from being stolen. People have to use your library (or write possibly worse algorithims from scratch) and that can hugely limit the amount of damage they can do with their closed-source, since the library and what it does is open, and because the license requires that reverse-engineering be allowed.
IMHO putting the GPL on libraries seems to be an attempt to lock the *interface* so that non-GPL code cannot call it. This sounds suspiciously like what MicroSoft does. It also does not work, the result (like for readline) is that nobody uses the library at all. Even for GPL code, since part of the GPL is the freedom to change your own code to be non-GPL if possible, and using that library means that is not possible.
Could somebody with Opera 6 please confirm or deny this idea? What is needed is evidence that this stylesheet somehow "fixes" Opera 6, like comparing screenshots of Opera 6 displaying this data and displaying the normal IE data.
Microsoft defenders keep saying "Opera 6 shows bullets misaligned" without actually putting out any proof that this stylesheet fixes those. From what little I know of css this can't possibly do so as it adjusts things with and without bullets equally.
Conversely it does not seem that Opera is willing to come up with a screenshot proving the opposite, that this page actually breaks Opera 6. It is possible that the sheet is useless but displays legibly in Opera 6 and thus could be a mistake by Microsoft that they would not detect unless they ran Opera 7. It would also be interesting to see how well IE's version displays in Opera 6 and whether it is bad enough that Microsoft felt they should try to fix it (and then could concievably have made a broken style sheet and given up at that point and forgotten about it).
You have posted this claim several times in direct contradiction to what Opera claims. If you want to prove this claim:
post a screenshot of Opera 6 displaying this page "correctly". Also post what Opera 6 does with the page delivered to IE.
Saying "opera 6 drew some bullets misaligned" proves NOTHING unless you show that this "broken" stylesheet actually fixes that problem, rather than making a worse display. Opera claims otherwise and has screen shots. Please prove your point.
If you bothered to read Opera's technical description you would see that "other" is sent the same stylesheet as IE, and Opera would have worked with that.
MSN specifically looks for Opera 7 and only it is sent the broken style sheet.
The broken style sheet would NEVER work, in any version of Opera or any other browser, at any time in history.
I'm sorry, it is almost impossible to come up with an explanation other than a direct attempt to break Opera.
The timeout before the focus change was the solution I first thought of and seems like the obvious one. However after thinking about it awhile (but not really having any system to test it on) I think the idea of waiting for a keystroke to change the menubar may be much better. A system that does not rely on a timeout would be much cleaner and predictable.
The other problem is something common to Linux and Windows, which is a confusion between being on top and having the focus. Unfortunately sometimes these are entangled quite badly at low levels in the system (X does not have this problem but new X window managers do). I can think of no other reason why Carbon has to be on top than the fact that the carbon interface to raise and give focus are the same and thus there is no way to tell the program otherwise. The double-buffered Quartz could easily emulate any graphic effects such as reading back from the screen so there is no technical reason why Carbon cannot act like Cocoa, it is certainly a limitation of the Carbon API.
Your problem with terminals is exactly what I have been yelling about here for months: CLICK SHOULD NOT RAISE WINDOWS!!!! This has been an endless frustration as it makes overlapping windows useless. And every time I say it some moron will say "but if some raise and some don't it will be inconsistent and confuse the user". That is bull and paranoid thinking like that is going to keep alternative GUI's from ever being better than Windows.
The first thing I thought of was to not have the menubar change for awhile so that moving the mouse up to it over other windows does not change it.
But now I think a much better idea is to not change it until you type to a window. Thus you can move the mouse to point at and highlight windows all you want, but the menubar does not change and thus you can continue moving to it to pull down a menu. Typing any key (including pushing down the command key for a menu shortcut) will switch the menubar to match the current focuesed window.
Clicking on the titlebar of a window would raise it, make it have the focus, and change the menubar.
Clicking inside a window would probably act exactly like the titlebar, if I can predict how Apple would design it. However I would prefer the apps to not raise unless the click was in a "dead" area such as blank background or anything where clicking does nothing, but clicking on a button or text field does not raise the window. In this case if you click on a button it is not clear if this should change the menubar. Anything that raises the window, including the app doing it itself, should change the menubar.
I don't find the "standardize the position of OK" argument convincing. This could also be "fixed" by moving the OK on a one-button dialog box left so it is in the same position as the two-button one.
The truth of all this argument is that OK-right was the design used on Macintoshes, which most post-Mac Unix software copied. There were some theoretical arguments for why this was better and thus selected by the Mac, but they are not really strong.
Like usual it was Microsoft that ignored prevailing standards and set their own and reversed the order (they also added the yes/no/cancel type dialogs, which had the annoying effect of reversing what yes/no meant when exiting a program compared to the Mac standard).
However imho the cause is lost. Microsoft set the standard and everybody (not just Gnome, but Mac) should give up and follow it. The alternatives do not have strong enough arguments for the this standard to be ignored.
A "tamper proof watermark" imho is easy to attain, if they keep the method of adding and removing the watermark secret, and do not provide any easy way of detecting if the watermark exists or has been removed
Unfortunately the music industry is stupid and will insist on making this somehow do copy prevention. That has the totally counter-productive result of providing a pirate with a fast and easy and foolproof method of determining if they have removed the watermark (ie if it plays they have succeeded in removing it).
As far as "deliberate attempt to destroy another's product", how about giving away IE for free:). Or that time they used a reserved field in the Kerberos protocol which broke compatibility with non-MS Kerberos software...
Even though both of those might be evil plans by Microsoft, both of them can also be explained as legitimate mistakes. The free IE can legitimately be said to make Windows a better product (ie it is better to be able to browse the web than to not be able to without installing something). The programmers of their Kerebos could very well have not realized that their use of an unassigned field meant that nobody else could work with their version.
But I still feel there is no reason for this Opera bug. I cannot imagine any possible scenario other than a deliberate attempt to "make Opera not work" for this to happen. Read the article, the document had no chance of working on *ANY* version of Opera (or any other browser), in any concievable way. And they explicitly checked for a very new version of Opera, not for some older version.
Unless Microsoft is purposly sending a broken page to IE2 this is NOT the same thing.
That can be explained as programmer error, or saying "we don't care if it does not work in IE2".
Sending html that only works in IE is done all the time by many sites, including Microsoft. Everybody knows about that and accepts it, grundginly, but it can be easily explained by laziness or lack of desire to fix it, not by deliberate malice.
Read the article. Microsoft deliberatly sends a broken page, when if they had done nothing (like a place that assummes everything is IE would do) then it would have displayed correctly.
This is one of the first things I have seen Microsoft do where there is absolutly no explanation other than deliberate attempt to destroy another's product.
Bill should not sign your check for such a stupid post. You have set back Microsoft shills quite a lot with inane stuff like this.
Read the article. The IE page displays CORRECTLY in Opera! There is no need to identify Opera and send a different page. And there is no browser in the world that displays the page sent to Opera correctly, including any version of Opera and any version of IE.
That would be a good solution for the user agent string in all the non-IE browsers. Have the "identify as this" be site-specific. That way you can set it only for the sites that are explicitly broken, but not mess up browser-usage counters or mess up sites that really truthfully are fixing a problem in your type of browser.
Re:Unbelievable what some people will claim!
on
Columbia Coverage
·
· Score: 1
Perhaps that is what it says. I didn't read it that way because that implies the test itself, which the article seemed to be praising, was actually useless. My reading was that "new foam worked" implies that "old foam did not work".
It could very well be that what the article really says is "this test we are talking about is useless as it failed to show any difference in the foams, while the actual shuttle missions prove there is a difference"
They may be covering things up, but I much more suspect the test really did prove something to their satisfaction, and they failed to explain clearly why this test was actually a success. Possibly they did alter the foam some more. It would be nice if they could prove that the test would reproduce the flaking by using some version of the foam.
You yourself said they voluntarily did not use freon (interestingly enough the "liberal" newspapers did not print this fact, leading to the much more damaging conclusion that it was an EPA *requirement*, not a Nasa decision, that could have caused the crash. So much for claims of liberal bias!). I suppose you could make an argument that the EPA requirements forced the company to change so much stuff to non-freon that it was cost-prohibitive to continue using freon for propellent in this one product.
While you are at it, it appears the shuttle does not have any asbestos in it (I looked after I saw how much freon was used), I bet you can make an argument that that caused the crash as well. Maybe it hit a protected heron during launch, and if we had just been allowed to shoot them all the shuttle would have been saved.
Re:Unbelievable what some people will claim!
on
Columbia Coverage
·
· Score: 1
And the old foam didn't flake, the new foam did. Read it again, or put in the appropriate terms into google and get other sources. I'm glad your comment will stand permanently on slashdot as proof that knee-jerk tree-huggers can't read.
Um, okay, I read it again:
"Initial results of the flight tests at Dryden, which were designed to replicate the pressure environment the Shuttle encounters in the first 65 seconds after launch, indicate the new foam survived the tests in perfect shape, with no evidence of flaking or erosion found."
Searching the web as you suggest reveals that huge amounts of freon are used in the shuttle, mostly in cooling and portions of the engines. Somehow your evil tree-huggers didn't stop it from being used elsewhere.
Didn't know that. Can you type//machine/c$foo to get c:foo on that machine? Can you type c$foo to get the same as c:foo on the local machine? Why didn't they use colon instead of dollar sign there?
Um, TurboTax IS proprietary software, so saying you have a desire to get rid of proprietary software but you won't do so until you can do so and run TurboTax is a contradiction.
A lot of people ask "can OSS produce as good of a word processor or CAD program or tax preparer" as commercial software? The answer may very well be no! (the reason OSS word processsors are close is because MicroSoft has no competition and thus no competitive drive any more, but in other fields it is obvious that OSS is nowhere near commercial offerings).
Most people take "no" to mean that Linux is doomed. Linux defenders will go to extreme ridiculous lengths to say some combination of OSS software will work kind of like the commercial software.
But what nobody realizes is that there is nothing wrong with running commercial software on Linux. Linux will win when commercial software is ported to it, not when people working for free duplicate all that commercial work. This has already happened in the visual effects industry, and in database and lots of imbedded uses. It can happen on the consumer desktop.
But it is not going to happen the way that both proponents and opponents of Linux claim, with all-free software.
The "average user" buys the machine at the store, plugs it in, and it boots up the operating system.
All this talk about ease of configuration or installation is meaningless. No "average user" is ever going to be expected to do what you did, with either Linux or Windows.
I do say that Linux is really stupid with the device names. Why did they copy the Windows drive letters being assigned in sequential order? They could have used the scsi id and disk partition number and other such information so that the same disk+partition had the same name no matter what other ones were added.
What are you talking about? Every single Linux program with a GUI I have ever seen has File/Save As in exactly the same place as Windows! It is EXACTLY the same!
Linux has quite a few problems, I think everybody here will admit, and it seems hard to believe you could not find a real one.
Slashdot Mirror
If you want to cut & paste the code so the same algorithim can be used by both your C program and shaders you can use macros. Ie "#define gl_vec2 vec2" can be inserted into the shader language fragment and then you can use gl_vec2 the way you want.
There is a world of difference between encryption and DRM. Encryption relies on two parties both of who are interested in keeping a secret. DRM attempts to make a party that is NOT interested in keeping a secret to do so. This violates the laws of physics and causality and is in the end impossible. Attempting to make the impossible possible will just make things incredibly inconvienent. The fact that it also makes competition with Microsoft impossible also makes everybody really pretty upset.
If you pay TrollTech money you can get a version you can use in closed-source programs. So it is still possible to write closed-source code.
Good one! The HDR code is available for commercial use if you negotiate a license with Digital Domain. Sort of like Qt. Also the algorithim is simple enough that you can probably recreate it without looking at the code.
I agree with all the joke responses that this is probably never going to happen in the USA.
The reason commercial software is statically-linked is that they are afraid of binary incompatabilities. Those static-linked libraries are the same ones.
It is obvious that the DNS server was written to allow closed-source software to call it. Otherwise it would be quite useless, people would not use it, or ignore the restriction (because it is impossible to enforce technically). I think all the really perverse examples that can be come up with it can be shown quite obviously that the writer intended non-GNU software to call it.
As for "libraries", it seems that all real-world examples are LGPL or similar. The only counter-example I have ever heard of is readline, can anybody (even RMS) name another? By this I mean a library that could in any concievable way be used by more than one program with significantly different functions. Libraries designed to work around the Windows DLL limitations so that "plugins" can load do not count.
In the real world (as opposed to RMS's fantasy wourld) the LGPL (or an even looser version that explicitly allows static linking, like I use) very well protects your clever algorithim from being stolen. People have to use your library (or write possibly worse algorithims from scratch) and that can hugely limit the amount of damage they can do with their closed-source, since the library and what it does is open, and because the license requires that reverse-engineering be allowed.
IMHO putting the GPL on libraries seems to be an attempt to lock the *interface* so that non-GPL code cannot call it. This sounds suspiciously like what MicroSoft does. It also does not work, the result (like for readline) is that nobody uses the library at all. Even for GPL code, since part of the GPL is the freedom to change your own code to be non-GPL if possible, and using that library means that is not possible.
Free software means the user can modify it, including modifying it to work around any hardware restriction.
Microsoft defenders keep saying "Opera 6 shows bullets misaligned" without actually putting out any proof that this stylesheet fixes those. From what little I know of css this can't possibly do so as it adjusts things with and without bullets equally.
Conversely it does not seem that Opera is willing to come up with a screenshot proving the opposite, that this page actually breaks Opera 6. It is possible that the sheet is useless but displays legibly in Opera 6 and thus could be a mistake by Microsoft that they would not detect unless they ran Opera 7. It would also be interesting to see how well IE's version displays in Opera 6 and whether it is bad enough that Microsoft felt they should try to fix it (and then could concievably have made a broken style sheet and given up at that point and forgotten about it).
post a screenshot of Opera 6 displaying this page "correctly". Also post what Opera 6 does with the page delivered to IE.
Saying "opera 6 drew some bullets misaligned" proves NOTHING unless you show that this "broken" stylesheet actually fixes that problem, rather than making a worse display. Opera claims otherwise and has screen shots. Please prove your point.
MSN specifically looks for Opera 7 and only it is sent the broken style sheet.
The broken style sheet would NEVER work, in any version of Opera or any other browser, at any time in history.
I'm sorry, it is almost impossible to come up with an explanation other than a direct attempt to break Opera.
The other problem is something common to Linux and Windows, which is a confusion between being on top and having the focus. Unfortunately sometimes these are entangled quite badly at low levels in the system (X does not have this problem but new X window managers do). I can think of no other reason why Carbon has to be on top than the fact that the carbon interface to raise and give focus are the same and thus there is no way to tell the program otherwise. The double-buffered Quartz could easily emulate any graphic effects such as reading back from the screen so there is no technical reason why Carbon cannot act like Cocoa, it is certainly a limitation of the Carbon API.
Your problem with terminals is exactly what I have been yelling about here for months: CLICK SHOULD NOT RAISE WINDOWS!!!! This has been an endless frustration as it makes overlapping windows useless. And every time I say it some moron will say "but if some raise and some don't it will be inconsistent and confuse the user". That is bull and paranoid thinking like that is going to keep alternative GUI's from ever being better than Windows.
The first thing I thought of was to not have the menubar change for awhile so that moving the mouse up to it over other windows does not change it.
But now I think a much better idea is to not change it until you type to a window. Thus you can move the mouse to point at and highlight windows all you want, but the menubar does not change and thus you can continue moving to it to pull down a menu. Typing any key (including pushing down the command key for a menu shortcut) will switch the menubar to match the current focuesed window.
Clicking on the titlebar of a window would raise it, make it have the focus, and change the menubar.
Clicking inside a window would probably act exactly like the titlebar, if I can predict how Apple would design it. However I would prefer the apps to not raise unless the click was in a "dead" area such as blank background or anything where clicking does nothing, but clicking on a button or text field does not raise the window. In this case if you click on a button it is not clear if this should change the menubar. Anything that raises the window, including the app doing it itself, should change the menubar.
The truth of all this argument is that OK-right was the design used on Macintoshes, which most post-Mac Unix software copied. There were some theoretical arguments for why this was better and thus selected by the Mac, but they are not really strong.
Like usual it was Microsoft that ignored prevailing standards and set their own and reversed the order (they also added the yes/no/cancel type dialogs, which had the annoying effect of reversing what yes/no meant when exiting a program compared to the Mac standard).
However imho the cause is lost. Microsoft set the standard and everybody (not just Gnome, but Mac) should give up and follow it. The alternatives do not have strong enough arguments for the this standard to be ignored.
Unfortunately the music industry is stupid and will insist on making this somehow do copy prevention. That has the totally counter-productive result of providing a pirate with a fast and easy and foolproof method of determining if they have removed the watermark (ie if it plays they have succeeded in removing it).
Even though both of those might be evil plans by Microsoft, both of them can also be explained as legitimate mistakes. The free IE can legitimately be said to make Windows a better product (ie it is better to be able to browse the web than to not be able to without installing something). The programmers of their Kerebos could very well have not realized that their use of an unassigned field meant that nobody else could work with their version.
But I still feel there is no reason for this Opera bug. I cannot imagine any possible scenario other than a deliberate attempt to "make Opera not work" for this to happen. Read the article, the document had no chance of working on *ANY* version of Opera (or any other browser), in any concievable way. And they explicitly checked for a very new version of Opera, not for some older version.
That can be explained as programmer error, or saying "we don't care if it does not work in IE2".
Sending html that only works in IE is done all the time by many sites, including Microsoft. Everybody knows about that and accepts it, grundginly, but it can be easily explained by laziness or lack of desire to fix it, not by deliberate malice.
Read the article. Microsoft deliberatly sends a broken page, when if they had done nothing (like a place that assummes everything is IE would do) then it would have displayed correctly.
This is one of the first things I have seen Microsoft do where there is absolutly no explanation other than deliberate attempt to destroy another's product.
Read the article. The IE page displays CORRECTLY in Opera! There is no need to identify Opera and send a different page. And there is no browser in the world that displays the page sent to Opera correctly, including any version of Opera and any version of IE.
That would be a good solution for the user agent string in all the non-IE browsers. Have the "identify as this" be site-specific. That way you can set it only for the sites that are explicitly broken, but not mess up browser-usage counters or mess up sites that really truthfully are fixing a problem in your type of browser.
It could very well be that what the article really says is "this test we are talking about is useless as it failed to show any difference in the foams, while the actual shuttle missions prove there is a difference"
They may be covering things up, but I much more suspect the test really did prove something to their satisfaction, and they failed to explain clearly why this test was actually a success. Possibly they did alter the foam some more. It would be nice if they could prove that the test would reproduce the flaking by using some version of the foam.
You yourself said they voluntarily did not use freon (interestingly enough the "liberal" newspapers did not print this fact, leading to the much more damaging conclusion that it was an EPA *requirement*, not a Nasa decision, that could have caused the crash. So much for claims of liberal bias!). I suppose you could make an argument that the EPA requirements forced the company to change so much stuff to non-freon that it was cost-prohibitive to continue using freon for propellent in this one product.
While you are at it, it appears the shuttle does not have any asbestos in it (I looked after I saw how much freon was used), I bet you can make an argument that that caused the crash as well. Maybe it hit a protected heron during launch, and if we had just been allowed to shoot them all the shuttle would have been saved.
Um, okay, I read it again:
"Initial results of the flight tests at Dryden, which were designed to replicate the pressure environment the Shuttle encounters in the first 65 seconds after launch, indicate the new foam survived the tests in perfect shape, with no evidence of flaking or erosion found."
Searching the web as you suggest reveals that huge amounts of freon are used in the shuttle, mostly in cooling and portions of the engines. Somehow your evil tree-huggers didn't stop it from being used elsewhere.
Didn't know that. Can you type //machine/c$foo to get c:foo on that machine? Can you type c$foo to get the same as c:foo on the local machine? Why didn't they use colon instead of dollar sign there?
A lot of people ask "can OSS produce as good of a word processor or CAD program or tax preparer" as commercial software? The answer may very well be no! (the reason OSS word processsors are close is because MicroSoft has no competition and thus no competitive drive any more, but in other fields it is obvious that OSS is nowhere near commercial offerings).
Most people take "no" to mean that Linux is doomed. Linux defenders will go to extreme ridiculous lengths to say some combination of OSS software will work kind of like the commercial software.
But what nobody realizes is that there is nothing wrong with running commercial software on Linux. Linux will win when commercial software is ported to it, not when people working for free duplicate all that commercial work. This has already happened in the visual effects industry, and in database and lots of imbedded uses. It can happen on the consumer desktop.
But it is not going to happen the way that both proponents and opponents of Linux claim, with all-free software.
All this talk about ease of configuration or installation is meaningless. No "average user" is ever going to be expected to do what you did, with either Linux or Windows.
I do say that Linux is really stupid with the device names. Why did they copy the Windows drive letters being assigned in sequential order? They could have used the scsi id and disk partition number and other such information so that the same disk+partition had the same name no matter what other ones were added.
Linux has quite a few problems, I think everybody here will admit, and it seems hard to believe you could not find a real one.