The C++ support in 2.95.2 is absurdly much better than 2.7. Yes 2.7 will compile lots of line noise vaguelly resembling C++, that 2.95.2 will rightfully reject. If you have code written using a "gcc 2.7 accept its, it must be good" mindset, you have my sympathy. You will never get a real C++ compiler to accept it.
GCC 2.95.2 is actually close to being a complete C++ compiler, the major lack is a standard conforming library. Yet, what it has is infinitely closer to the standard than the old libg++ bundled with GCC 2.7. And don't even think avbout using STL with GCC 2.7.
The major thing holding back GCC 3.0 is completing the rewrite of the C++ standard library, which is *not* included with Red Hat GCC 2.96.
It has _always_ been Sun's strategy to concentrate their effort on as few platforms (hardware and OS) as feasible. Sun is quicker than the competition to drop support for old platforms. Sun has been pretty succesful in their strategy, why change it now?
Remember that Red Hat these days are also Cygnus, who have always had their own commercially supported gcc releases in parallel with the official FSF.
I have no doubt this new Red Hat released gcc is at least as good as the official FSF released gcc the other distributers bundle. It is probably even better. Which in itself is scary for several reasons:
1. Calling it gcc 2.96 is certain to cause confusion, as evident by this thread.
2. It gived Red Hat an advantage over the other Linux distributors, which could be seen as unfair.
3. It might encourage other distributors to include gcc snapshots in order to seem competitive. But they don't have the same Cygnus derived competance, so they are more likely to screw up, and different gcc snapshot, maybe all called gcc 2.96, and maybe with binary incompatible C++ libraries, is bound to create chaos.
So I'd have prefered that Red Hat bundled the FSF released gcc as the default compiler, and had the "Cygnus" compiler as a different product (pehaps also bundled, but clearly separate, and not default).
Gee, I thought it was a real nerd joke. The X10 Window System and the X-Men. Like the T-shirts that came out at the same time as the Malcolm X movie, featuring a huge X, and in smaller print "version 11, release 4" and in even smaller print "don't worry, It's a nerd thing".
Now what is this _other_ X10 thing? They have an particular ugly web side, which made it difficult to see what they were trying to sell.
For MacOS using sprintf to create the string to draw with QuickDraw would open up the same sort of problems as printf does with console (or MPW!) applications.
Some information is missing here. A license doen't need to be compatible with the GPL for debian to software covered by it. It just has to be free. Debian includes lots of free software covered by licenses that are not compatible with the GPL.
Will some Debian comment what the real issue is?
PS: I predict we will see lots of flames here (condemning RMS, Debian, Python or the GPL) from people who haven't bothered to find out what the issue really is.
4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt
otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance.
This is the exact clause from the GPL which RMS is talking about. There is no such clause in the old BSD license, so there is no need for any legal "forgiveness" after breaking the old BSL license.
That was your first error. Your second error was to assume that there was a license conflict in the first place. Since glibc is covered by LGPL and not GPL, there is no conflict with the old BSD license.
Your third error is to assume there is a need for an apology in case of a conflict. There isn't. There is (at most) a need for a legal forgiveness, to counteract section four of the GPL. An apology will have no legal significance.
I agree that the need for a legal "forgiveness" for distributing GPL'ed code linked with Qt is absurd. But is it the law that is absurd, or RMS interpretation of the law?
Probably a bit of both, I can't see any logic flaws in RMS's interpretation, but he has always been much more strict in following copyright law to the letter than just about everyone else, where other people ignore problems that would never occur in practice.
Sometimes this pays off, like when he insisted that the GNU project developed its own patent free compression program (gzip), while just about everyone else was satisfied with using compress and LZW, and trusted the vague statements from Unisys that they would never enforce their patent on software only systems.
What is wrong to do is to take this as an insult to KDE. Being paranoid about copyright law is a fundamental part of RMS's nature.
When Cygnus "took control" of gcc from RMS, it was for very good reason. RMS had not been competently maintaining the source tree, and
many would-be code contributors who had good solutions to genuine problems were getting increasingly disgusted.
RMS had not been maintaining GCC at all. He had stoped working on GCC and left maintenance to Richard Kenner. Everybody involved with GCC (EGCS) agrees RK is an excellent compiler engineer, but meny felt he wasn't a good release engineer. EGCS was created as an experiment, to see if a different form would attract more outside contributions. This was done in agreement with RMS. The experiment was a success, EGCS became the officiel GCC, and RK now work as a compiler engineer on the new GCC team.
Re:FSF now prefers Qt/KDE over Gtk/Gnome
on
Qt Going GPL
·
· Score: 1
> But this is all speculation, isn't it?
Yes, but you are right. The stated reasons for prefering GPL over LGPL doesn't really apply in this case. However, the idea of FSF advocating Qt over Gtk+ is fun.
Re:FSF now prefers Qt/KDE over Gtk/Gnome
on
Qt Going GPL
·
· Score: 2
> They sold a proprietrary version of gcc to
> Motorola (which they could only do because they
> require copyright to be signed over to them).
At least anonymous/. posters claim this over and over, despite having no supporting evidence, and despite the FSF having signed hundreds of counter contracts with GCC contributers stating that they will not allow proprietary version of GCC.
Re:Red Hat covets non-free software?
on
Qt Going GPL
·
· Score: 2
It wasn't meant as an attack on Red Hat, I believe attracting unfree software is good for Linux. For example, it means I can play some cool games under Linux, which would probably not be released under a free software license.
Using the LGPL rather than the GPL for key libraries, is one way to attract unfree software.
QPL is still an option
on
Qt Going GPL
·
· Score: 2
which mean you can release your software under any free software license you want.
If you want to make your software unfree, you won't get Qt for free either. Basically, you choose the rules (free or non-free), and they will apply to both your software and Qt. What you cannot choose, is to play by different rules for Qt and your own software. Somehow, this fair to me.
I don't know a lot of this stuff, but the "easy" port would be to Cygwin + Xlib, and using one of the existing win32 X servers. I doubt this would compete with Troll Techs own native Win32 port.
If you want to make a native win32 port of Qt Free Edition, Cygwin would probably not be of help.
Re:Port XEmacs to KDE!
on
Qt Going GPL
·
· Score: 2
> Go Per!:-) Just kidding. I love your work man.
Actually, I suspect Bill Perry (of Emacs/W3 fame) wish this had happened three months ago. He has just finished porting XEmacs to gtk/gnome, which he chose because of licensing reasons, even though he prefered Qt for technical reasons.
Annoying for us who develop free software that must run on both Unix and MS Windows. However, with the Unix version going GPL, it is now our own fault if we don't port it to win32.
> This is still a move to praise, but with
> the understanding that the qt/gtk competition
> is still important.
Yes, to developers of unfree software, and to companies like Red Hat which want to attract unfree software to Linux. However, I suspect most free software developers don't see it as a problem, that developers of unfree software will have to pay Troll Tech for a license to the professional version.
If it ain't broke, don't fix it!
on
Qt Going GPL
·
· Score: 2
> Why is it companies never open source stuff
> when things are going their way?
It would be silly to change a business model that *worked*. Only when the currect business model breaks down, you go looking for alternatives.
Or do you believe companies should make free software for the good of their hearts?
Another win for freedom of choice!
on
Qt Going GPL
·
· Score: 3
> I thought OSS was about *freedom*, freedom
> of choice.
Well, Qt 2.2 gives the developer the *choice* of two licenses, QPL or GPL. Qt 2.0 only offered one license, the QPL. So it seems to me that the freedom of choice has increased, not decreased, with this announcement.
FSF now prefers Qt/KDE over Gtk/Gnome
on
Qt Going GPL
·
· Score: 4
Well, I have seen no announcement to that effect. But given that FSF prefer libraries to be covered by the GPL rather than the LGPL, they should now consider Qt preferable over Gtk.
Since companies like Red Hat have a interest in promoting development af Linux software, free or unfree, they should still have an interest in Gtk, because the LGPL allows their customers to develop unfree applications.
> I'm wondering just how pervasive the US
> computer caffine cluture is internationally.
Seen from Denmark the coffein obsession is an American phenomen. Everybody here drink some disgustingly strong tasting coffee in absurd quantities, and the idea of decaffenated anything (coffee, tea, cola) is meet with puzzlement. Why would anyone want that? The point being, most people here are coffein junkies, and don't know it or recognize it as a problem.
Nerds are probably *less* coffein-addicted than the average population, simply because cola contains much less coffein than coffee.
Re:Paid to solve problems
on
Men of Zeal
·
· Score: 2
Mozilla (average wage is around US$ 100.000 p.a., I thinks that is competitive... it is certainly a lot more than I make). The money come primarily from Internet access fees.
Re:Paid to solve problems
on
Men of Zeal
·
· Score: 2
Well, you sound like a "Man of Zeal". I consider free software a development model that should be used where it works better than traditional models, and not be used where it is worse. I believe the market will sort out which areas are in which category.
Personally, I'm happy to get my salary, solve my employers solutions, and contribute to free software, all at the same time. I consider a solution that offers that a full solution for me. Apparently, you require the economy must work exactly like it does for traditional development in order to be a full solution for you.
However, your last claim, that in order for all software to be free the development *must* be financed exactly like traditional development, is clearly false. Appart from the "helping to sell unfree software" part, any of your suggestions would ensure that software will continue to be produced in a free software world, and combined they will ensure that a lot of free software is produced, whether or not there exists unfree software as well.
So I would suggest that you started accepting other form for financing than those of the traditional development form. All your suggested forms involve getting payed for doing good work that your customers are happy for and that benefits others, so the only problems are those you create in your own head.
Slashdot Mirror
The C++ support in 2.95.2 is absurdly much better than 2.7. Yes 2.7 will compile lots of line noise vaguelly resembling C++, that 2.95.2 will rightfully reject. If you have code written using a "gcc 2.7 accept its, it must be good" mindset, you have my sympathy. You will never get a real C++ compiler to accept it.
GCC 2.95.2 is actually close to being a complete C++ compiler, the major lack is a standard conforming library. Yet, what it has is infinitely closer to the standard than the old libg++ bundled with GCC 2.7. And don't even think avbout using STL with GCC 2.7.
The major thing holding back GCC 3.0 is completing the rewrite of the C++ standard library, which is *not* included with Red Hat GCC 2.96.
It has _always_ been Sun's strategy to concentrate their effort on as few platforms (hardware and OS) as feasible. Sun is quicker than the competition to drop support for old platforms. Sun has been pretty succesful in their strategy, why change it now?
Remember that Red Hat these days are also Cygnus, who have always had their own commercially supported gcc releases in parallel with the official FSF.
I have no doubt this new Red Hat released gcc is at least as good as the official FSF released gcc the other distributers bundle. It is probably even better. Which in itself is scary for several reasons:
1. Calling it gcc 2.96 is certain to cause confusion, as evident by this thread.
2. It gived Red Hat an advantage over the other Linux distributors, which could be seen as unfair.
3. It might encourage other distributors to include gcc snapshots in order to seem competitive. But they don't have the same Cygnus derived competance, so they are more likely to screw up, and different gcc snapshot, maybe all called gcc 2.96, and maybe with binary incompatible C++ libraries, is bound to create chaos.
So I'd have prefered that Red Hat bundled the FSF released gcc as the default compiler, and had the "Cygnus" compiler as a different product (pehaps also bundled, but clearly separate, and not default).
Gee, I thought it was a real nerd joke. The X10 Window System and the X-Men. Like the T-shirts that came out at the same time as the Malcolm X movie, featuring a huge X, and in smaller print "version 11, release 4" and in even smaller print "don't worry, It's a nerd thing".
Now what is this _other_ X10 thing? They have an particular ugly web side, which made it difficult to see what they were trying to sell.
For MacOS using sprintf to create the string to draw with QuickDraw would open up the same sort of problems as printf does with console (or MPW!) applications.
Some information is missing here. A license doen't need to be compatible with the GPL for debian to software covered by it. It just has to be free. Debian includes lots of free software covered by licenses that are not compatible with the GPL.
Will some Debian comment what the real issue is?
PS: I predict we will see lots of flames here (condemning RMS, Debian, Python or the GPL) from people who haven't bothered to find out what the issue really is.
That was your first error. Your second error was to assume that there was a license conflict in the first place. Since glibc is covered by LGPL and not GPL, there is no conflict with the old BSD license.
Your third error is to assume there is a need for an apology in case of a conflict. There isn't. There is (at most) a need for a legal forgiveness, to counteract section four of the GPL. An apology will have no legal significance.
I agree that the need for a legal "forgiveness" for distributing GPL'ed code linked with Qt is absurd. But is it the law that is absurd, or RMS interpretation of the law?
Probably a bit of both, I can't see any logic flaws in RMS's interpretation, but he has always been much more strict in following copyright law to the letter than just about everyone else, where other people ignore problems that would never occur in practice.
Sometimes this pays off, like when he insisted that the GNU project developed its own patent free compression program (gzip), while just about everyone else was satisfied with using compress and LZW, and trusted the vague statements from Unisys that they would never enforce their patent on software only systems.
What is wrong to do is to take this as an insult to KDE. Being paranoid about copyright law is a fundamental part of RMS's nature.
> In addition he talks about KDE specifically and
> says nothing about other Qt apps so your point
> is moot.
You really should read the article before commenting. The stated "forgiveness" explicitly covers al linking with Qt, KDE isn't mentioned in that part.
> There is no code copyrighted by the FSF in KDE.
> Not a single line.
Maybe, but the legal "forgiveness" includes all applications that links FSF owned code with Qt. Are you saying there exists no such applications?
> But this is all speculation, isn't it?
Yes, but you are right. The stated reasons for prefering GPL over LGPL doesn't really apply in this case. However, the idea of FSF advocating Qt over Gtk+ is fun.
> They sold a proprietrary version of gcc to
/. posters claim this over and over, despite having no supporting evidence, and despite the FSF having signed hundreds of counter contracts with GCC contributers stating that they will not allow proprietary version of GCC.
> Motorola (which they could only do because they
> require copyright to be signed over to them).
At least anonymous
It wasn't meant as an attack on Red Hat, I believe attracting unfree software is good for Linux. For example, it means I can play some cool games under Linux, which would probably not be released under a free software license.
Using the LGPL rather than the GPL for key libraries, is one way to attract unfree software.
which mean you can release your software under any free software license you want.
If you want to make your software unfree, you won't get Qt for free either. Basically, you choose the rules (free or non-free), and they will apply to both your software and Qt. What you cannot choose, is to play by different rules for Qt and your own software. Somehow, this fair to me.
I don't know a lot of this stuff, but the "easy" port would be to Cygwin + Xlib, and using one of the existing win32 X servers. I doubt this would compete with Troll Techs own native Win32 port.
If you want to make a native win32 port of Qt Free Edition, Cygwin would probably not be of help.
> Go Per! :-) Just kidding. I love your work man.
Actually, I suspect Bill Perry (of Emacs/W3 fame) wish this had happened three months ago. He has just finished porting XEmacs to gtk/gnome, which he chose because of licensing reasons, even though he prefered Qt for technical reasons.
Annoying for us who develop free software that must run on both Unix and MS Windows. However, with the Unix version going GPL, it is now our own fault if we don't port it to win32.
> This is still a move to praise, but with
> the understanding that the qt/gtk competition
> is still important.
Yes, to developers of unfree software, and to companies like Red Hat which want to attract unfree software to Linux. However, I suspect most free software developers don't see it as a problem, that developers of unfree software will have to pay Troll Tech for a license to the professional version.
> Why is it companies never open source stuff
> when things are going their way?
It would be silly to change a business model that *worked*. Only when the currect business model breaks down, you go looking for alternatives.
Or do you believe companies should make free software for the good of their hearts?
> I thought OSS was about *freedom*, freedom
> of choice.
Well, Qt 2.2 gives the developer the *choice* of two licenses, QPL or GPL. Qt 2.0 only offered one license, the QPL. So it seems to me that the freedom of choice has increased, not decreased, with this announcement.
Well, I have seen no announcement to that effect. But given that FSF prefer libraries to be covered by the GPL rather than the LGPL, they should now consider Qt preferable over Gtk.
Since companies like Red Hat have a interest in promoting development af Linux software, free or unfree, they should still have an interest in Gtk, because the LGPL allows their customers to develop unfree applications.
> I'm wondering just how pervasive the US
> computer caffine cluture is internationally.
Seen from Denmark the coffein obsession is an American phenomen. Everybody here drink some disgustingly strong tasting coffee in absurd quantities, and the idea of decaffenated anything (coffee, tea, cola) is meet with puzzlement. Why would anyone want that? The point being, most people here are coffein junkies, and don't know it or recognize it as a problem.
Nerds are probably *less* coffein-addicted than the average population, simply because cola contains much less coffein than coffee.
Mozilla (average wage is around US$ 100.000 p.a., I thinks that is competitive... it is certainly a lot more than I make). The money come primarily from Internet access fees.
Well, you sound like a "Man of Zeal". I consider free software a development model that should be used where it works better than traditional models, and not be used where it is worse. I believe the market will sort out which areas are in which category.
Personally, I'm happy to get my salary, solve my employers solutions, and contribute to free software, all at the same time. I consider a solution that offers that a full solution for me. Apparently, you require the economy must work exactly like it does for traditional development in order to be a full solution for you.
However, your last claim, that in order for all software to be free the development *must* be financed exactly like traditional development, is clearly false. Appart from the "helping to sell unfree software" part, any of your suggestions would ensure that software will continue to be produced in a free software world, and combined they will ensure that a lot of free software is produced, whether or not there exists unfree software as well.
So I would suggest that you started accepting other form for financing than those of the traditional development form. All your suggested forms involve getting payed for doing good work that your customers are happy for and that benefits others, so the only problems are those you create in your own head.