No. You have the right to remain silent when you are under arrest or are no longer free to leave. But even beyond that, your right to remain silent is to allow you to prevent incriminating yourself. Police do not have to mirandize you if 1) the information is non-incriminating in nature or 2) the information you've given is given voluntarily. Giving your name says nothing about your guilt or innocence. This is what the justices have ruled, as indicated in their holdings:
But if an officer does not have to say what you are being stopped on suspicion of, how are you to know whether your name is or is not incriminating?
It seems to, in effect, nullify the Fifth Amendment with respect to your identity.
In Canada the police in fact can be compelled to tell you why you are being arrested.
They can in the US as well.
The problem is that this ruling allows them to demand information *without* "intending" to arrest you.
One of the justifications that the Supreme Court listed is that allowing this behavior helps police officers find people with outstanding warrants. I just think that we have enough problems with police on fishing expeditions without bypassing Miranda and the Fifth Amendment to facilitate the same thing.
When Blizzard *does* go under and Battle.net service stops being offered, I'm sure that everyone will appreciate the presence of bnetd, despite Blizzard's attempts to squash it.
The problem is that the suspect asked whether he was being arrested, and if so, why. Since this was a Terry stop, he did not recieve Miranda rights protection.
The problem is that invoking the Fifth Amendment requires knowing that self-incrimination is possible (which is why Hiibel's argument of Fifth Amendment protection was rejected by the US Supreme Court, as he shouldn't have been worried about self-incrimination). There is no real way to *know* whether you are at risk of self-incrimination without a police officer disclosing what they are considering charging you with. This basically renders useless Fifth Amendment protection against releasing identity, even though the US Supreme Court specifically said that the Fifth Amendment *could* apply to releasing one's identity.
This is a severely broken system. If police have no reason to potentially charge someone, they have no reason to stop them. If they have such a reason, I do not understand why they cannot be compelled to inform the person of what they are being accused of.
The kernel isn't a library, though. I think that the closest problem that could come up would be non-GPL compatible modules using kernel functionality. Whether or not he is legally correct, Linus has stated that he doesn't consider GPL-incompatible modules to be infringing.
Use of a GPL library, like Qt, entails that people not produce GPL-incompatible software.
Yes, but Microsoft purchased Connectix and acquired Virtual PC, which has an emulation engine that should be fast enough to emulate the x86 processor in the XB1 at the XB2's processor speed.
Presumably the appeal of PowerPC is that, while emulating x86 on PowerPC is somewhat feasible (if a bit slow) because of the much greater number of PPC general-purpose registers, it is impossible to emulate the PowerPC on the x86 without severe performance impacts.
This is a very nice copy protection method for Microsoft -- all those Windows users out there can't just download an XB2 emulator and start playing.
You know, I really like seeing Ars-Slashdot ties. I don't see the two as competing much -- Ars has lots of good, original content, whereas Slashdot just links to content, but Slashcode is (IMHO) a more pleasant-to-use forum engine. They complement each other very nicely.
Well, yes, but proprietary is a very large umbrella term -- it can refer to everything from the Windows source code to the patents Apple holds on TrueType kerning. I meant a way of distinguishing between software that is free-as-in-beer to use and distribute, where source can be downloaded and learned from with no NDAs or registrations or anything being signed and the rest of the body of what we generally would call "non-Free", "non-OSI-compliant" software.
They would be wanting to use millions of man-hours of work without contributing back to the pool, and I don't think that's a reasonable position to adopt.
It is certainly less efficient, but not all software of this class is like the NPL, where the original author of the work gets special rights to improvements. Povray or xmame, for instance, have non-OSI-compliant licenses, but do not grant special rights to anyone. I guess that I can agree that it might be rather nice if everyone had intercompatible licenses, but I don't agree that the GPL is the "ultimate" license or that it's a good idea to try to force all GUI desktop software for the premier hobbyist OS to fall under a particular license (or pay a penalty fee for not doing so). I'm just not comfortable with it.
Povray's kind of a disappointment. I completely understand why they can't release it under an open license, and I certainly don't hold those reasons against them, but it's sad that the (best? only?) source-available raytracer still has so many strings attached. Here's to hoping that they can get it all resolved.
There are a number of GPL-compatible raytracers. It's not really all that hard to write a raytracer -- I did one for a class project once. YafRay is GPLed, CoolRay is GPLed, and Blender has an integrated raytracer (as well as other renderers).
Not really. It would leave Qt stranded as GPL-only, with not even the remotest possibility of creating an open source application, even if you had the money.
It's a standard procedure on/. Where's the fun in mod points if you can't reward the people who agree with you and punish the ones who don't
Maybe, but I've criticized Linux issues before, argued that Palladium/TCPA are weak and ineffectual and not particularly dangerous, and Apple issues on apple.slashdot.org and not been modded as flamebait for doing so -- I've seen many +5s. I mean, if Slashdot was literally nothing but groupthink, I wouldn't spend my time here.
Did you have any other questions? You seem to have a pretty tenuous grasp on what constitutes an Open Source license. I strongly suggest you read the definition before you make wild and incorrect claims about what it means.
Hmm...you're right. I always thought that the OSI defined "Open Source" and that "open source" predated them and referred to source code public availability, but it appears that I was wrong.
Well, there does not appear to be a concise name for the class of software that I am discussing -- source free-as-in-beer-available and free-as-in-beer usable but not OSI-compliant. I do consider that these packages are a not insignificant chunk of the world, though, and they are certainly used and more-poorly-served by Qt's license scheme than the license schemes that the FSF uses or Linus uses.
FWIW, povray theoretically is being completely rewritten for v4 to be open source, according to discussion I've read, but I've seen little indication of this actually happening -- it's a pretty large task.
Trolltech could never dominate KDE or Linux. 'Cause if they would start doing that, KDE can easely fork Qt and continue on it's own.
It is certainly true that, if KDE is willing to become X11-only, it would be possible for the KDE project to fork Qt at any point. However, it would also force KDE and Qt and all software that uses it to forever be GPL-compatible (excuse me -- there was a correction further down -- OSI-accepted licenses would also be accepted).
Even if Trolltech would ceise to exist there is no problem, because the current Qt would then become public under a BSD-liscense.
Right. However, I do not view that agreement as a sufficient degree of security. I brought up that point in my post:
If they want to, they can increase fees on Qt to whatever they desire (fees based on expected return of a product using Qt, for instance), they can place whatever demands on companies licensing the non-GPL Qt that they'd like (as long as they continue to offer the non-free Qt under some terms, the KDE agreement protections do not kick in. For whoever thinks that companies will never turn harmful, I give the example of Caldera becoming SCO).
If you think that I'm under some misapprehensions concerning Qt licensing scheme, you're welcome to clear those up, to argue that some concerns are not of issue in the real world, or to correct mistakes.
It is certainly not a flamebait post, though I admit to having strong worries about the use of Qt. Saying "GNOME is teh suck" or "KDE blows and the developers are stupid" would be flamebait, with no content. This might have errors, it might contain emotional bias, and it might make some people unhappy, but it is not flamebait.
Fair enough -- I should include OSI-approved but non-GPL-compatible software. This is not remotely all OSS software: consider pine, povray, XFree86, anything under the old BSD license, qmail, djbdns, etc.
Have you ever used pine, povray, XFree86, qmail, or xv? All these (and a host more) are major open source software packages that are not GPL-compatible. TrollTech wants to hand a partial solution out in the hopes that it will give them complete control over the Linux GUI world.
"Discontinue" is a lovely loophole for TrollTech. Any company would take that kind of control over, say Win32 in a *moment*. Microsoft doesn't even have that kind of power.
If they're doing closed source non-GPLed development
Whoah, careful on combining things. There's a lot of open source software out there that *isn't* GPL-compatible, believe it or not.
There are also other issues some folks (including me) have with Qt's licensing situation.
You can't use GPLed libraries in proprietary software without some other licensing arrangement. Didn't we know this already?
Consider the fact that none of the other fundamental libraries for a typical Linux system (libc, libm, libpthread, libpam, libutil, libdl, libcrypt, libncurses, libz, libX11, libXt, libXext, libSM, libICE, libGL) are GPLed. This was not by accident. It was to ensure that Linux is a free-as-in-beer development environment for *all* users.
This is not some strike against Qt, it's just more FUD (probably from a GNOME zealot).
I think I'm reasonably unbiased -- I don't even use the GNOME desktop environment, though I do prefer to use GTK-based applications.
Sure, the GNOME libraries are LGPL. They still suck donkey balls compared with KDE's development platform.
You have a very liberal interpretation of "sucking donkey balls". It's subjective of course -- I have my own peeves with Qt, like use of a weird preprocessor and disregard for the STL, and could with about as much justification (well, more, since I listed two issues) say the same thing about Qt.
You are using hyperbole -- the Qt licensing problems are real, serious, not likely to go away, and brushed off by the Ars Technica article. I've mentioned a few of the problems that I have with Qt further down. It is wildly different from using a GPLed kernel (the GPL largely does not enforce issues between the kernel and userspace apps, and Linus has taken an interpretation that the GPL is not forced upon kernel modules). BSD software does not force software using it to take a particular license, and the library at issue is not one of the most fundamental to Linux's future use.
I think that the article dodged around a number of the license problems with Qt.
Among other things:
* Qt is GPLed, and as such, prevents FOSS developers from using any OSS licenses that are not compatible with the GPL. Ironically enough, XFree86's license is not compatible with the GPL, and hence XFree86 could not include a Qt configuration utility.
* Ideological issues. Many developers and users move to Linux to avoid being under the control of a single company -- Qt is an attempt to change all that.
* TrollTech makes their money by hurting Linux. TrollTech has established themselves in a position where they can collect a tax on a vast number of developers *if* Qt becomes accepted as the standard Linux widget set. There is no such tax to develop non-GPL-compatible software for Windows -- I can get a copy of mingw or cygwin and, without paying a cent, sit down and write a closed-source application or an GPL-incompatible license for Windows. In that respect, Windows is more free than Linux is. Try telling a shareware author that instead of being able to write software freely, as he could for years on Windows and MacOS, that on *Linux* he must pay an up-front fee of $1500 *per* person that has contributed a line of code to the project for every piece of GUI software he writes. Wait and see his reaction. Even Stallman, the strongest GPL advocate in existence, does not expect everyone on a platform to be forced to use the GPL -- glibc is LGPLed, and the GPL does not attempt to force everyone on a *platform* to use the GPL. Readline (which is GPL) is not a valid comparison -- there are alternatives to readline. To maintain consistenct behavior in a GUI environment, *everyone* must use the same GUI widget set.
* TrollTech has seized a phenomenal amount of power over the future of Linux. If they want to, they can increase fees on Qt to whatever they desire (fees based on expected return of a product using Qt, for instance), they can place whatever demands on companies licensing the non-GPL Qt that they'd like (as long as they continue to offer the non-free Qt under some terms, the KDE agreement protections do not kick in. For whoever thinks that companies will never turn harmful, I give the example of Caldera becoming SCO). The omission of anything guaranteeing user and developer security is, IMHO, unlikely to be purely innocent -- this is a document that was carefully gone over by lawyers. TrollTech has firmly ensconced themselves in a position to exploit the entire future GUI environment on Linux. Remember that Linus Torvalds and many other people license their software specifically under the GPLv2, because they do not trust the FSF not to abuse their position and revise the GPL in such a way to allow wielding a body of IP that has never had an equal. In a similar way, many of us do not trust TrollTech, an organization founded for the sole purpose to make money (unlike the more ideologically-oriented FSF) not to take the simple step of revising their terms of sale on the QPL to exploit a phenomenal body of IP. Time upon time in the past, when a company has a knife to the throat of other companies in the software development community and times turn harsh or stockholders want a bigger return, that knife is used. I'd rather not go through a massive, bloody situation on Linux. GTK is just too good of an alternative, and it's too easy to avoid the entire mess.
* It is not feasible to do C Qt programs. This is a technical issue rather than a legal one. Even Microsoft did Win32 and then built MFC on top of it, allowing folks to use C if they want to. Qt imposes C++ no matter what one does (or a high-level language like Python calling C++). This breaks a long-standing *IX tradition of not *requiring* anyone to use anything but C, if a developer so desires. GTK+, on the other hand, has a C API and a standard set of C++ bindings built on top, a more conventionally accepted approach.
I'd like to reiterate that there are a lot of people that like KDE that m
It's really making me sad how many people are excitedly awaiting the features IE "will have in SP2 or Longhorn".
I'm really kind of dubious that anyone is, in fact "excitedly awaiting IE features". The kind of people that get excited over web browser features are already using Gecko-based browsers.
Slashdot Mirror
No. You have the right to remain silent when you are under arrest or are no longer free to leave. But even beyond that, your right to remain silent is to allow you to prevent incriminating yourself. Police do not have to mirandize you if 1) the information is non-incriminating in nature or 2) the information you've given is given voluntarily. Giving your name says nothing about your guilt or innocence. This is what the justices have ruled, as indicated in their holdings:
But if an officer does not have to say what you are being stopped on suspicion of, how are you to know whether your name is or is not incriminating?
It seems to, in effect, nullify the Fifth Amendment with respect to your identity.
In Canada the police in fact can be compelled to tell you why you are being arrested.
They can in the US as well.
The problem is that this ruling allows them to demand information *without* "intending" to arrest you.
One of the justifications that the Supreme Court listed is that allowing this behavior helps police officers find people with outstanding warrants. I just think that we have enough problems with police on fishing expeditions without bypassing Miranda and the Fifth Amendment to facilitate the same thing.
When Blizzard *does* go under and Battle.net service stops being offered, I'm sure that everyone will appreciate the presence of bnetd, despite Blizzard's attempts to squash it.
The problem is that the suspect asked whether he was being arrested, and if so, why. Since this was a Terry stop, he did not recieve Miranda rights protection.
The problem is that invoking the Fifth Amendment requires knowing that self-incrimination is possible (which is why Hiibel's argument of Fifth Amendment protection was rejected by the US Supreme Court, as he shouldn't have been worried about self-incrimination). There is no real way to *know* whether you are at risk of self-incrimination without a police officer disclosing what they are considering charging you with. This basically renders useless Fifth Amendment protection against releasing identity, even though the US Supreme Court specifically said that the Fifth Amendment *could* apply to releasing one's identity.
This is a severely broken system. If police have no reason to potentially charge someone, they have no reason to stop them. If they have such a reason, I do not understand why they cannot be compelled to inform the person of what they are being accused of.
The kernel isn't a library, though. I think that the closest problem that could come up would be non-GPL compatible modules using kernel functionality. Whether or not he is legally correct, Linus has stated that he doesn't consider GPL-incompatible modules to be infringing.
Use of a GPL library, like Qt, entails that people not produce GPL-incompatible software.
Sure, but you can't emulate a hard drive in ram (unless they're using flash ram, which I kinda doubt).
I'd be guessing that they'd rely on higher-capacity memcards taking up the slack.
LGPL code cannot be closed a la BSD, but it doesn't restrict people's choice of license in writing an application that uses it.
You couldn't take an LGPLed TCP stake, but you could use an LGPLed libjpeg.
Yes, but Microsoft purchased Connectix and acquired Virtual PC, which has an emulation engine that should be fast enough to emulate the x86 processor in the XB1 at the XB2's processor speed.
Presumably the appeal of PowerPC is that, while emulating x86 on PowerPC is somewhat feasible (if a bit slow) because of the much greater number of PPC general-purpose registers, it is impossible to emulate the PowerPC on the x86 without severe performance impacts.
This is a very nice copy protection method for Microsoft -- all those Windows users out there can't just download an XB2 emulator and start playing.
You know, I really like seeing Ars-Slashdot ties. I don't see the two as competing much -- Ars has lots of good, original content, whereas Slashdot just links to content, but Slashcode is (IMHO) a more pleasant-to-use forum engine. They complement each other very nicely.
"Propietary", and I'm not trying to be flippant.
Well, yes, but proprietary is a very large umbrella term -- it can refer to everything from the Windows source code to the patents Apple holds on TrueType kerning. I meant a way of distinguishing between software that is free-as-in-beer to use and distribute, where source can be downloaded and learned from with no NDAs or registrations or anything being signed and the rest of the body of what we generally would call "non-Free", "non-OSI-compliant" software.
They would be wanting to use millions of man-hours of work without contributing back to the pool, and I don't think that's a reasonable position to adopt.
It is certainly less efficient, but not all software of this class is like the NPL, where the original author of the work gets special rights to improvements. Povray or xmame, for instance, have non-OSI-compliant licenses, but do not grant special rights to anyone. I guess that I can agree that it might be rather nice if everyone had intercompatible licenses, but I don't agree that the GPL is the "ultimate" license or that it's a good idea to try to force all GUI desktop software for the premier hobbyist OS to fall under a particular license (or pay a penalty fee for not doing so). I'm just not comfortable with it.
Povray's kind of a disappointment. I completely understand why they can't release it under an open license, and I certainly don't hold those reasons against them, but it's sad that the (best? only?) source-available raytracer still has so many strings attached. Here's to hoping that they can get it all resolved.
There are a number of GPL-compatible raytracers. It's not really all that hard to write a raytracer -- I did one for a class project once. YafRay is GPLed, CoolRay is GPLed, and Blender has an integrated raytracer (as well as other renderers).
Not really. It would leave Qt stranded as GPL-only, with not even the remotest possibility of creating an open source application, even if you had the money.
:-)
I think you mean "closed-source".
It's a standard procedure on /. Where's the fun in mod points if you can't reward the people who agree with you and punish the ones who don't
Maybe, but I've criticized Linux issues before, argued that Palladium/TCPA are weak and ineffectual and not particularly dangerous, and Apple issues on apple.slashdot.org and not been modded as flamebait for doing so -- I've seen many +5s. I mean, if Slashdot was literally nothing but groupthink, I wouldn't spend my time here.
Well, who knows.
Did you have any other questions? You seem to have a pretty tenuous grasp on what constitutes an Open Source license. I strongly suggest you read the definition before you make wild and incorrect claims about what it means.
Hmm...you're right. I always thought that the OSI defined "Open Source" and that "open source" predated them and referred to source code public availability, but it appears that I was wrong.
Well, there does not appear to be a concise name for the class of software that I am discussing -- source free-as-in-beer-available and free-as-in-beer usable but not OSI-compliant. I do consider that these packages are a not insignificant chunk of the world, though, and they are certainly used and more-poorly-served by Qt's license scheme than the license schemes that the FSF uses or Linus uses.
FWIW, povray theoretically is being completely rewritten for v4 to be open source, according to discussion I've read, but I've seen little indication of this actually happening -- it's a pretty large task.
Trolltech could never dominate KDE or Linux. 'Cause if they would start doing that, KDE can easely fork Qt and continue on it's own.
It is certainly true that, if KDE is willing to become X11-only, it would be possible for the KDE project to fork Qt at any point. However, it would also force KDE and Qt and all software that uses it to forever be GPL-compatible (excuse me -- there was a correction further down -- OSI-accepted licenses would also be accepted).
Even if Trolltech would ceise to exist there is no problem, because the current Qt would then become public under a BSD-liscense.
Right. However, I do not view that agreement as a sufficient degree of security. I brought up that point in my post:
If they want to, they can increase fees on Qt to whatever they desire (fees based on expected return of a product using Qt, for instance), they can place whatever demands on companies licensing the non-GPL Qt that they'd like (as long as they continue to offer the non-free Qt under some terms, the KDE agreement protections do not kick in. For whoever thinks that companies will never turn harmful, I give the example of Caldera becoming SCO).
If you think that I'm under some misapprehensions concerning Qt licensing scheme, you're welcome to clear those up, to argue that some concerns are not of issue in the real world, or to correct mistakes.
It is certainly not a flamebait post, though I admit to having strong worries about the use of Qt. Saying "GNOME is teh suck" or "KDE blows and the developers are stupid" would be flamebait, with no content. This might have errors, it might contain emotional bias, and it might make some people unhappy, but it is not flamebait.
Excuse me -- I left a sentence unfinished:
Many people see the alternative to Qt becoming the standard GUI API as
this should read:
Many people see the alternative to Qt becoming the standard GUI API as a C API, like GTK+, being adopted, and are not thrilled at such a prospect.
Your first statement is wrong.
Fair enough -- I should include OSI-approved but non-GPL-compatible software. This is not remotely all OSS software: consider pine, povray, XFree86, anything under the old BSD license, qmail, djbdns, etc.
Have you ever used pine, povray, XFree86, qmail, or xv? All these (and a host more) are major open source software packages that are not GPL-compatible. TrollTech wants to hand a partial solution out in the hopes that it will give them complete control over the Linux GUI world.
"Discontinue" is a lovely loophole for TrollTech. Any company would take that kind of control over, say Win32 in a *moment*. Microsoft doesn't even have that kind of power.
If they're doing closed source non-GPLed development
Whoah, careful on combining things. There's a lot of open source software out there that *isn't* GPL-compatible, believe it or not.
There are also other issues some folks (including me) have with Qt's licensing situation.
You can't use GPLed libraries in proprietary software without some other licensing arrangement. Didn't we know this already?
Consider the fact that none of the other fundamental libraries for a typical Linux system (libc, libm, libpthread, libpam, libutil, libdl, libcrypt, libncurses, libz, libX11, libXt, libXext, libSM, libICE, libGL) are GPLed. This was not by accident. It was to ensure that Linux is a free-as-in-beer development environment for *all* users.
This is not some strike against Qt, it's just more FUD (probably from a GNOME zealot).
I think I'm reasonably unbiased -- I don't even use the GNOME desktop environment, though I do prefer to use GTK-based applications.
Sure, the GNOME libraries are LGPL. They still suck donkey balls compared with KDE's development platform.
You have a very liberal interpretation of "sucking donkey balls". It's subjective of course -- I have my own peeves with Qt, like use of a weird preprocessor and disregard for the STL, and could with about as much justification (well, more, since I listed two issues) say the same thing about Qt.
You are using hyperbole -- the Qt licensing problems are real, serious, not likely to go away, and brushed off by the Ars Technica article. I've mentioned a few of the problems that I have with Qt further down. It is wildly different from using a GPLed kernel (the GPL largely does not enforce issues between the kernel and userspace apps, and Linus has taken an interpretation that the GPL is not forced upon kernel modules). BSD software does not force software using it to take a particular license, and the library at issue is not one of the most fundamental to Linux's future use.
I think that the article dodged around a number of the license problems with Qt.
Among other things:
* Qt is GPLed, and as such, prevents FOSS developers from using any OSS licenses that are not compatible with the GPL. Ironically enough, XFree86's license is not compatible with the GPL, and hence XFree86 could not include a Qt configuration utility.
* Ideological issues. Many developers and users move to Linux to avoid being under the control of a single company -- Qt is an attempt to change all that.
* TrollTech makes their money by hurting Linux. TrollTech has established themselves in a position where they can collect a tax on a vast number of developers *if* Qt becomes accepted as the standard Linux widget set. There is no such tax to develop non-GPL-compatible software for Windows -- I can get a copy of mingw or cygwin and, without paying a cent, sit down and write a closed-source application or an GPL-incompatible license for Windows. In that respect, Windows is more free than Linux is. Try telling a shareware author that instead of being able to write software freely, as he could for years on Windows and MacOS, that on *Linux* he must pay an up-front fee of $1500 *per* person that has contributed a line of code to the project for every piece of GUI software he writes. Wait and see his reaction. Even Stallman, the strongest GPL advocate in existence, does not expect everyone on a platform to be forced to use the GPL -- glibc is LGPLed, and the GPL does not attempt to force everyone on a *platform* to use the GPL. Readline (which is GPL) is not a valid comparison -- there are alternatives to readline. To maintain consistenct behavior in a GUI environment, *everyone* must use the same GUI widget set.
* TrollTech has seized a phenomenal amount of power over the future of Linux. If they want to, they can increase fees on Qt to whatever they desire (fees based on expected return of a product using Qt, for instance), they can place whatever demands on companies licensing the non-GPL Qt that they'd like (as long as they continue to offer the non-free Qt under some terms, the KDE agreement protections do not kick in. For whoever thinks that companies will never turn harmful, I give the example of Caldera becoming SCO). The omission of anything guaranteeing user and developer security is, IMHO, unlikely to be purely innocent -- this is a document that was carefully gone over by lawyers. TrollTech has firmly ensconced themselves in a position to exploit the entire future GUI environment on Linux. Remember that Linus Torvalds and many other people license their software specifically under the GPLv2, because they do not trust the FSF not to abuse their position and revise the GPL in such a way to allow wielding a body of IP that has never had an equal. In a similar way, many of us do not trust TrollTech, an organization founded for the sole purpose to make money (unlike the more ideologically-oriented FSF) not to take the simple step of revising their terms of sale on the QPL to exploit a phenomenal body of IP. Time upon time in the past, when a company has a knife to the throat of other companies in the software development community and times turn harsh or stockholders want a bigger return, that knife is used. I'd rather not go through a massive, bloody situation on Linux. GTK is just too good of an alternative, and it's too easy to avoid the entire mess.
* It is not feasible to do C Qt programs. This is a technical issue rather than a legal one. Even Microsoft did Win32 and then built MFC on top of it, allowing folks to use C if they want to. Qt imposes C++ no matter what one does (or a high-level language like Python calling C++). This breaks a long-standing *IX tradition of not *requiring* anyone to use anything but C, if a developer so desires. GTK+, on the other hand, has a C API and a standard set of C++ bindings built on top, a more conventionally accepted approach.
I'd like to reiterate that there are a lot of people that like KDE that m
GNOME is a much more distributed project than KDE is.
I think that the approach to documentation, software design, policy and politics all reflect this.
It's really making me sad how many people are excitedly awaiting the features IE "will have in SP2 or Longhorn".
I'm really kind of dubious that anyone is, in fact "excitedly awaiting IE features". The kind of people that get excited over web browser features are already using Gecko-based browsers.
Or maybe it's a backup plan -- if you can't successfully sue everyone that has used Linux, sue everyone that has used Minix! :-)