Now you *know* what will happen. This will rear it's ugly head again whenever some GPLed piece of software based on Python gets popular...y'know, Python ain't elisp.:^)
There's a clause in the license of the Linux kernel. This apparently makes it all hunky-dorey.
BTW I was wrong about Nautilus--and this exposes a potential problem with the LGPL. It's not OK for QPL to give permission to link GPL software to QPL software, but it's OK to write LGPL wrappers to link to GPL software (which is how it's OK for GNOME software to use Mozilla.)
Hell, aparently there's a line in the GPL about not adding any further restrictions to the GPL (I don't know where it is, sorry) which the QPL did. The only restriction the QPL added is that you couldn't charge for distributing software (which the GPL, surprisingly enough, allows.)
It's the one thing that you'll never get out of an FSF zealot: just what the hell is wrong with the QPL?
You bring up an interesting point: apparently, it's OK for Nautilus to use Mozilla as a rendering engine, since it (Nautilus) calls Mozilla through Bonobo, which is LGPL. Ugh. Truthfully, the KDE team should have seen this...except that zealots always argue against this strategy (except when it works to their advantage.) I'm not sure about using the LGPL to circumvent licensing problems--to my way of thinking, this is *more* of a problem than KDE not being fully GPL ever was.
BTW Thanks to whoever pointed out how wrong I originally was about the Mozilla/Nautilus debate--sorry, I was sleepy when I came up with the idea, and ran with it.:^) It wasn't the first time I went on a stupid rant, and I'm sure it won't be the last.:^)
The former statement was written in ignorance of how exactly Mozilla was embedded.
However, it brings up an intriguing possibility-the KDE license dispute could have been settled long ago! The Harmony project could have simply had someone write up a spec (which I believe was in the works at one time) and could have written an LGPL wrapper lib to cover up the licensing issue. No fuss, no muss.
Now, the former paragraph is clearly wrong--that wouldn't solve the problem. Or would it? Apparently, it does for Nautilus--the interface between Mozilla and Nautilus has been Bonoboized. What does that mean? Mozilla is called through Bonobo, which is LGPLed, which in turn uses CORBA to communicate--and, as it has been explained to me, this means that there is *no linking* of Mozilla to Nautilus.
Is this true? Is this all that is required? We just use the LGPL library to do as we see fit? Wow...you know, this doesn't seem right, somehow. It wasn't OK for the QPL to give permission to link not-for-charge GPL programs to GPLed programs, but it's OK to write an LGPL wrapper for anything--and link to GPLed software with not legal problems.
Yet another reason I would never consider the GPL/LGPL...
Naw, thanks to the piss-poor way the GPL is written, a good lawyer could state that a Python script is a derivative work of Python (after all, you need Python to run it, right? And it's not a system essential like libc or GCC, right?) And your script is GPL? Gee, here comes the GPL Police.
The important thing to remember is that many supposed violations of the GPL are really just abuses of the GPL (which are easy since the GPL is so poorly worded.)
The problem isn't merely his attacks on various GPLed (and non-GPLed) projects, but his selectivity. He seems, at times, to only go after what he doesn't like (or doesn't support to begin with.)
For example, up until a week ago, he went after KDE like a pirhanna after a monkey in the water. The sad thing was that the problem was with the GPL restricting one's freedom to link (the ambiguous "derivative work" clause), not the QPL. Even sadder is the fact that he's *not* pursuing people linking to Mozilla within the GNOME project.
Come on, RMS, either you persue license violations or you don't...
The problem with visiting the FSF site is that, in many cases, the site offers no *proof.* I want proof, not just bold statements to the effect of "This license is incompatible with the GPL and therefore shouldn't be used to write free software."
Thank you for wasting your time and everyone elses time, and, oh yeah, space on Slashdot's server and the bandwidth it took to post that. Thanks a fucking lot.
>[How] is this a VMWare competitor? Just because
>Linux users use VMWare to run windows?
Bingo, ya jackass! That's exactly the point. Don't be stupid.:^)
>Does Win4Lin let you install other OS's on it?
>Run multiple concurrent VMs?
Hrm, with a name like Win4Lin, what do you think your chances of running BeOS on it are?
God, I hate dumb people...and I hate smartasses. You seem to know *exactly why* this was used as a competitor to VMWare, but pretend to not. Why? Do you think it will do some kind of good? I mean, come on, it was listed as a competitor because a lot of folks (I used to use it for this reason) use VMWare to run Windows. If all you need to do is run Windows (the 95/98 variety) then this is an option to choose other than VMWare...making them competitors.
Still don't get it? Both products potentially allow one to run Windows on a virtual machine in Linux. Win4Lin is a more focused, more finely targeted product. Win4Lin doesn't have all the bells and whistles of VMWare, but that's not the point. The point is that both run Windows on a VM under Linux.
> I could be completly wronge here, but if it
>does indeed use MS code, and runs in kernel
>space, i can't imagine this setup being very
>stable.
Yeah, but if you want to be able to actually run Windows under Linux, you need Ring 0--which means hooking into the kernel.
And yes, I have Win4Lin and yes, it's rock solid.:^) I'm guessing this has to do with their hardware virtualisation, and the way it's implemented. I haven't managed to crash it yet.
>Another point to make is that this is a comercial >product. (way) not GPL.
Boo-hoo. I assume you're reading this comment under either Lynx, W3M or Emacs, because you sure as hell can't be using Netscape under X11R6, since neither of those are GPLed.
If you don't like the fact that it's GPLed, why not help with either plex86 or help get Dosemu to work in protected mode? Warning--both projects will probably end up hooking into the kernel, so they won't be stable at all.;^)
If people would bother to think things through, they would realize that many folks look to VMWare and Win4Lin for the same reasons--to run Windows under Linux. When looking at the two products in that way, Win4Lin becomes *very* attractive.
If you're an OS developer, or you want to be able to run Win95, 98, NT, and BeOS (BTW, if you run more than one copy of Windows at at time you violate the EULA. Read it some time.) then VMWare is the way to go. If you just need Win95/98, run Win4Lin.
How did this thread of argument get started, anyway? It's the stupidest fuckin' argument I've ever seen.
Uh, from the standpoint of someone who merely wants to run Windoze apps under Linux, they might as well be the same.:^)
NO, NO, NO Win4Lin is NOT "like" Wine! Wine provides an API to run Win16/32 apps on top of, while Win4Lin provides a layer to run Windows on. The reason for the kernel modifications, BTW, are probably for running in protected mode (which is why Dosemu can't run Win95/98/ME.)
I've talked to a number of folks who use VMWare basically to run Windows, and that's kinda how VMWare markets its product, so, yeah, it's a fair comparison.
While the focus of Win4Lin is different, and the implementation is different, if all one wants is Windows apps without rebooting, it's a fair comparison.
Actually, my original post (if there was doubt) was meant to be sarcastic...I was commenting on the fact that the "illegality" in question stems from a portion of the GPL pertaining to derivative works. Some have used that to state that both linking to Qt and borrowing from other, non-Qt-linked GPL programs are both illegal.
As far as I can tell, it only states that one cannot use GPLed code without permission of the original author. OK, fair enough...anyone know of any other GPLed software guilty of this heinous crime? Quite frankly, all I ever heard about was KDE software. Any takers?
And, quite frankly, I saw the GPL'ing of Qt as being a sad day for free software. There was *no* problem with the QPL. The QPL gave explicit permission for GPL programs to be linked to QPLed software. The problem, friends, was with the GPL. Do we fix it? No...it helped support the FSF's agenda of the day, and it shall remain vague and ambiguous for the same reason.
Thank you, FSF, for getting mixed up in (bad) politics.
> He is defending *your* right 'to choose' and your right of 'freedom of speech'.
*My* problem with some of RMSs statements is that he's forgiving KDE and Troll Tech for what was, in fact, a problem with a badly-worded (and much in need of rewording) section of the the GPL.
/*
He couldn't quite manage to argue that the QPL was non-free, but he did manage to argue that it was
incompatible with the GPL (I doubt this claim would stand up in court)
See this site on this very topic.
*/
I did.
And guess what? There's just a claim that the license is incompatible, along with a disclaimer that can be added to the COPYING file that somehow magically makes it all compatible.
The real issue: an ambiguous reading of the GPL that claims that dynamic linking is a derivative work. It's a dubious claim, and one that most certainly wouldn't win in a court of law (although IANAL) without a lot of legal posturing. That Troll Tech caved in is proof that they were merely tired of hearing the whining. I mean, come on, they offered the toolkit free for free (as in beer AND speech) software because they benefitted from the existence of free software. As I recall, their original dev platform was a Linux box.:^)
In short, a simple claim of incompatibility isn't a real claim at all--it's just bullshit in the disguise of a real claim. RMS just used his position of Saint IGNUcius (which is the most offensive thing he's ever done, IMHO) to dictate what is and what is not incompatible.
oh brother--I'm tired of hearint the "would have hurt hardware sales" argument. The problem was Motorola, and their inability to turn out enough processors. That's still the problem today.
Why would this have been OK for their hardware sales? They could have used x86 processors and built their own motherboards, graphics hardware, etc., and not have given anyone else their specs. True, someone could have stolen the specs and made their own, but remember the example of the Laser 128 built by Franklin--don't mess with Apple in a court of law.:^)
Slashdot Mirror
Now you *know* what will happen. This will rear it's ugly head again whenever some GPLed piece of software based on Python gets popular...y'know, Python ain't elisp. :^)
There's a clause in the license of the Linux kernel. This apparently makes it all hunky-dorey.
BTW I was wrong about Nautilus--and this exposes a potential problem with the LGPL. It's not OK for QPL to give permission to link GPL software to QPL software, but it's OK to write LGPL wrappers to link to GPL software (which is how it's OK for GNOME software to use Mozilla.)
Hell, aparently there's a line in the GPL about not adding any further restrictions to the GPL (I don't know where it is, sorry) which the QPL did. The only restriction the QPL added is that you couldn't charge for distributing software (which the GPL, surprisingly enough, allows.)
:^) It wasn't the first time I went on a stupid rant, and I'm sure it won't be the last. :^)
It's the one thing that you'll never get out of an FSF zealot: just what the hell is wrong with the QPL?
You bring up an interesting point: apparently, it's OK for Nautilus to use Mozilla as a rendering engine, since it (Nautilus) calls Mozilla through Bonobo, which is LGPL. Ugh. Truthfully, the KDE team should have seen this...except that zealots always argue against this strategy (except when it works to their advantage.) I'm not sure about using the LGPL to circumvent licensing problems--to my way of thinking, this is *more* of a problem than KDE not being fully GPL ever was.
BTW Thanks to whoever pointed out how wrong I originally was about the Mozilla/Nautilus debate--sorry, I was sleepy when I came up with the idea, and ran with it.
The former statement was written in ignorance of how exactly Mozilla was embedded.
However, it brings up an intriguing possibility-the KDE license dispute could have been settled long ago! The Harmony project could have simply had someone write up a spec (which I believe was in the works at one time) and could have written an LGPL wrapper lib to cover up the licensing issue. No fuss, no muss.
Now, the former paragraph is clearly wrong--that wouldn't solve the problem. Or would it? Apparently, it does for Nautilus--the interface between Mozilla and Nautilus has been Bonoboized. What does that mean? Mozilla is called through Bonobo, which is LGPLed, which in turn uses CORBA to communicate--and, as it has been explained to me, this means that there is *no linking* of Mozilla to Nautilus.
Is this true? Is this all that is required? We just use the LGPL library to do as we see fit? Wow...you know, this doesn't seem right, somehow. It wasn't OK for the QPL to give permission to link not-for-charge GPL programs to GPLed programs, but it's OK to write an LGPL wrapper for anything--and link to GPLed software with not legal problems.
Yet another reason I would never consider the GPL/LGPL...
Neither Nautilus nor Galeon can ever become an "official" part of GNOME, since Mozilla is currently released under an incompatible license.
What goes for one, goes for everything else.
Naw, thanks to the piss-poor way the GPL is written, a good lawyer could state that a Python script is a derivative work of Python (after all, you need Python to run it, right? And it's not a system essential like libc or GCC, right?) And your script is GPL? Gee, here comes the GPL Police.
The important thing to remember is that many supposed violations of the GPL are really just abuses of the GPL (which are easy since the GPL is so poorly worded.)
>KDE not complying is damn important, it
>undermines everything that does. Python doesn't? >Well it should and it will.
I agree..and Nautilus and Galeon should too.
Nautlus && Galeon == GPL
Mozilla == MPL
According to FSF, MPL & GPL are incompatible
Therefore Nautilus && Galeon == illegal
The problem isn't merely his attacks on various GPLed (and non-GPLed) projects, but his selectivity. He seems, at times, to only go after what he doesn't like (or doesn't support to begin with.)
For example, up until a week ago, he went after KDE like a pirhanna after a monkey in the water. The sad thing was that the problem was with the GPL restricting one's freedom to link (the ambiguous "derivative work" clause), not the QPL. Even sadder is the fact that he's *not* pursuing people linking to Mozilla within the GNOME project.
Come on, RMS, either you persue license violations or you don't...
The problem with visiting the FSF site is that, in many cases, the site offers no *proof.* I want proof, not just bold statements to the effect of "This license is incompatible with the GPL and therefore shouldn't be used to write free software."
Do NOT try to go into an honest business distributing music in the MP3 format. It will only get you sued into oblivion. Instead, trade MP3s illegally.
Or so this case would make it seem.
Thank you for wasting your time and everyone elses time, and, oh yeah, space on Slashdot's server and the bandwidth it took to post that. Thanks a fucking lot.
Jerk.
How about:
*Photoshop
*QuarkXPress
*Exchange client
I fail to find Linux versions of these. Could you point me in the right direction?
>[How] is this a VMWare competitor? Just because
:^)
>Linux users use VMWare to run windows?
Bingo, ya jackass! That's exactly the point. Don't be stupid.
>Does Win4Lin let you install other OS's on it?
>Run multiple concurrent VMs?
Hrm, with a name like Win4Lin, what do you think your chances of running BeOS on it are?
God, I hate dumb people...and I hate smartasses. You seem to know *exactly why* this was used as a competitor to VMWare, but pretend to not. Why? Do you think it will do some kind of good? I mean, come on, it was listed as a competitor because a lot of folks (I used to use it for this reason) use VMWare to run Windows. If all you need to do is run Windows (the 95/98 variety) then this is an option to choose other than VMWare...making them competitors.
Still don't get it? Both products potentially allow one to run Windows on a virtual machine in Linux. Win4Lin is a more focused, more finely targeted product. Win4Lin doesn't have all the bells and whistles of VMWare, but that's not the point. The point is that both run Windows on a VM under Linux.
> I could be completly wronge here, but if it
:^) I'm guessing this has to do with their hardware virtualisation, and the way it's implemented. I haven't managed to crash it yet.
;^)
>does indeed use MS code, and runs in kernel
>space, i can't imagine this setup being very
>stable.
Yeah, but if you want to be able to actually run Windows under Linux, you need Ring 0--which means hooking into the kernel.
And yes, I have Win4Lin and yes, it's rock solid.
>Another point to make is that this is a comercial >product. (way) not GPL.
Boo-hoo. I assume you're reading this comment under either Lynx, W3M or Emacs, because you sure as hell can't be using Netscape under X11R6, since neither of those are GPLed.
If you don't like the fact that it's GPLed, why not help with either plex86 or help get Dosemu to work in protected mode? Warning--both projects will probably end up hooking into the kernel, so they won't be stable at all.
If people would bother to think things through, they would realize that many folks look to VMWare and Win4Lin for the same reasons--to run Windows under Linux. When looking at the two products in that way, Win4Lin becomes *very* attractive.
If you're an OS developer, or you want to be able to run Win95, 98, NT, and BeOS (BTW, if you run more than one copy of Windows at at time you violate the EULA. Read it some time.) then VMWare is the way to go. If you just need Win95/98, run Win4Lin.
How did this thread of argument get started, anyway? It's the stupidest fuckin' argument I've ever seen.
Uh, from the standpoint of someone who merely wants to run Windoze apps under Linux, they might as well be the same. :^)
NO, NO, NO Win4Lin is NOT "like" Wine! Wine provides an API to run Win16/32 apps on top of, while Win4Lin provides a layer to run Windows on. The reason for the kernel modifications, BTW, are probably for running in protected mode (which is why Dosemu can't run Win95/98/ME.)
I've talked to a number of folks who use VMWare basically to run Windows, and that's kinda how VMWare markets its product, so, yeah, it's a fair comparison.
While the focus of Win4Lin is different, and the implementation is different, if all one wants is Windows apps without rebooting, it's a fair comparison.
That's another good point.
Actually, my original post (if there was doubt) was meant to be sarcastic...I was commenting on the fact that the "illegality" in question stems from a portion of the GPL pertaining to derivative works. Some have used that to state that both linking to Qt and borrowing from other, non-Qt-linked GPL programs are both illegal.
As far as I can tell, it only states that one cannot use GPLed code without permission of the original author. OK, fair enough...anyone know of any other GPLed software guilty of this heinous crime? Quite frankly, all I ever heard about was KDE software. Any takers?
And, quite frankly, I saw the GPL'ing of Qt as being a sad day for free software. There was *no* problem with the QPL. The QPL gave explicit permission for GPL programs to be linked to QPLed software. The problem, friends, was with the GPL. Do we fix it? No...it helped support the FSF's agenda of the day, and it shall remain vague and ambiguous for the same reason.
Thank you, FSF, for getting mixed up in (bad) politics.
Thank you for forgiving Troll Tech and KDE for committing the sin of violating an ambiguity in the GPL.
>What is the problem ?
> He is defending *your* right 'to choose' and your right of 'freedom of speech'.
*My* problem with some of RMSs statements is that he's forgiving KDE and Troll Tech for what was, in fact, a problem with a badly-worded (and much in need of rewording) section of the the GPL.
/*
:^)
He couldn't quite manage to argue that the QPL was non-free, but he did manage to argue that it was
incompatible with the GPL (I doubt this claim would stand up in court)
See this site on this very topic.
*/
I did.
And guess what? There's just a claim that the license is incompatible, along with a disclaimer that can be added to the COPYING file that somehow magically makes it all compatible.
The real issue: an ambiguous reading of the GPL that claims that dynamic linking is a derivative work. It's a dubious claim, and one that most certainly wouldn't win in a court of law (although IANAL) without a lot of legal posturing. That Troll Tech caved in is proof that they were merely tired of hearing the whining. I mean, come on, they offered the toolkit free for free (as in beer AND speech) software because they benefitted from the existence of free software. As I recall, their original dev platform was a Linux box.
In short, a simple claim of incompatibility isn't a real claim at all--it's just bullshit in the disguise of a real claim. RMS just used his position of Saint IGNUcius (which is the most offensive thing he's ever done, IMHO) to dictate what is and what is not incompatible.
Bastard.
What are you talking about?
:^)
Are you talking about the Macmillan deal? Macmillan ships Red Hat GPL, too. Seriously, please post a reply, because I'm confused as hell.
Having said that: oh great, I just upgraded from 7.0 to 7.1.
oh brother--I'm tired of hearint the "would have hurt hardware sales" argument. The problem was Motorola, and their inability to turn out enough processors. That's still the problem today.
:^)
Why would this have been OK for their hardware sales? They could have used x86 processors and built their own motherboards, graphics hardware, etc., and not have given anyone else their specs. True, someone could have stolen the specs and made their own, but remember the example of the Laser 128 built by Franklin--don't mess with Apple in a court of law.
Hrm, they were probably thinking, "Gee, it really sucks that Motorola can't make all the PowerPC chips we need...Intel x86's are cheap...why not?"
:^)
BTW, that's the same problem they're having today. That's why they're talking about it *again.* Stay tuned.
As long as they add a clause to the GPL, it'll be fine.
If you believe the folks at Debian, that is.