You didn't read what you posted carefully enough. Go back, read it again. It simply states you *must* cause any work derive from the Program or any part thereof to be licensed under the License(GPL). That's why Sun's committed a violation. They need to either do what you suggest they've already done, or stop their distribution.
I used to see this bit used as an argument against (the formerly non-GPL) Qt. That was a good argument, too, except that it's wrong. It only pertains to a program released under this License or to code derived from the Program already under this License. True, the argument was also that the QPL wasn't GPL compatible, but, c'mon, Troll Tech had given their *permission* in their FAQ.
Of course, this is a dead issue even though RMS won't let it drop.:^)
What about the fact that the drivers are a derived work of the kernel? The COPYING file that comes with the kernel states that using the headers in a user-level program aren't a derived work, true, but surely there's a problem with Bruce's argument.
Surely some sort of code within the kernel has to be used in writing the driver. I cannot say with authority since I have never written a kernel driver. But I'm hoping that this post will spur someone to respond to this issue.
Shit, that beats the hell out of the new kidneys I bought, and the brain. The new brain works almost as good as the old one, but I'm having some troble with my logic/reasoning. Oh well, a couple more years, and cracking open my old logic book, and I'll be right as rain *tic*
You know, there will be plenty of flames (esp. on Slashdot) about this article, but there's one thing to keep in mind:
He's absolutely right.
ESR's description of how Linux was developed isn't really accurate. Many folks already knew it (or, like myself, have read it in other sources.:^) Linux is held together with excellent management skills. I've even seen statements from ESR to that effect.
One could state that this is merely semantics, but it's not entirely correct. It's true that people were not part of some corporation like IBM or Compaq; they were mostly just people who wanted/needed a feature/fix for the kernel, and did it. But that kernel growth/improvement was *managed*, not just some big cloud of code that condensed on a server somewhere.
Absolutely true (and I'm guilty of not having read the article yet.)
Simply stating that it's not true that a project isn't self-organizing is true...and it's false. If there were no interest in the project, and there as a fair amount of dissent between the project coordinator and those who wanted to contribute, the project would die. In a sense, it really *is* self-organizing. IMHO people like Linux are mostly like a managing editor is for a magazine or newspaper--some guiding vision with lots of damage control.
It's not a thing like a traditional company--in a traditional company, people are working towards a common goal so they can get a paycheck. In an open-source project, people are working (hopefully) towards a common goal so they'll have something they'll be able to use.
Really??? That's SWEEEEEEEEEEEEEEEET!!!!:^) This would allow the sort of thing I've been concerned with...at least, I think it will.:^) I'll have to do a re-read of the QPL, but I was wondering if it allowed greater freedom of what you can and cannot use as a license for your "derived work.":^)
Read again--I made no such claim. It was *I* who wished that Troll Tech would release under the LGPL rather than the GPL. I never said that Troll Tech wanted or wished that they could release Qt under the LGPL. By releasing under the GPL rather than the LGPL, they've just made things more difficult for folks who might choose to release software based on Free QT--LGPL makes some of those licensing difficulties magically melt away.
Perhaps the Troll Tech crew is hoping to derive code out of Qt later for use in a non-library. The LGPL won't allow that (check out the LGPL--derived works *must* be a library AFAIK). If that's the case then, well, OK.
>it is always seemed to me that the GNU Public >License exhibits an intolerance to even the >slightest differences in philosophy.
Absolutely. There is no tolerance for difference of opinion in the GPL. And if you do a careful reading of the GPL, you'll learn just how *ambiguous* the GPL is, even though it seems to be highly specific. I think the best example is "derivative work."
My reasoning? Yeah, it seems pretty clear that the FSF had a pretty good claim. But the GPL takes the beauty of linked libraries and just stomps all over it. The whole point of a library is that you can have this pre-compiled piece of code and have it work with what you're compiling right now. Did you have to have the source to compile your code? Aside from headers, no. Where's the wording about headers? Er, um...huh. How about dynamic linking? Is the object code of your program compiled with offending object code? Yes and no.
Ooh, here's a case that'll leave you wondering. Suppose I release what is, for all appearances, a source- and binary- compatible GTK clone. Then I release it under a license that doesn't require me to release source, and I distribute it as binary-only. Users install it, because I've made a much lighter, much faster GTK. They run Gnome apps on this. Who's the violator? Me? No, I hired someone to write a specification for me, and based my code on the spec, not GTK source. The GNOME team? Nope...they want nothing to do with me. The user? Well, wait; mine's binary-compatible with GTK. All the user had to do was drop in my libs to an already working system. Where's the violation? There's one somewhere; where is it?
Even better: let's say some company purchases a commercial license for Qt and releases a dynamically-linked version of their closed-source commercial app. I install the app on my system and run it; after all, I installed the Qt RPMs.:^) Am I in trouble here? Maybe. But's a pretty ambiguous situation.
>1) You just imagine a hypothetical situation,
> draw your conclusions about Mr. Stallman's
> reaction and then want us to blame him about
> this reaction YOU imagined.
I believe the author was referring to the QPL/GPL debate, and wrapping it in a thinly-veiled hypothetical situation.
AFAIK the problem pertained to the fact that the QPL was subject to the laws of Norway, and that the GPL doesn't allow that. That wouldn't have been a problem if someone hadn't gotten half-witted and decided that dynamically-linked libraries were derivative works. This means, of course, that I cannot write software for free operating systems, as I'm under the constant fear that, not wanting to release software under the GPL, that any/all authors who have released libs under the LGPL will switch to the GPL (which the FSF advocates.)
Oh well, I've wanted to learn about MFC for a while now.:^)
But it was a misinterpretation of this statement that caused certain Debian members to go off half-cocked and state that KDE was illegal.
If anyone should be begging for forgiveness, it's the debian team (or at least the offending(offensive?)) members. Now Troll Tech has been forced into releasing Qt under the GPL (NOT the LGPL, as they should have) due to the constant whining (oh, you don't think they were forced? I think they're just trying to save face) and we now have less choice in what license a KDE app can be released under. Want to release something that meets Troll Tech's "free" requirements, yet keep it open in case you want to "go commercial" later? Well, TOO FRICKIN' BAD!
PS Please prove the last statement wrong...it would make me feel better about the world.:^)
is that Qt switching from QPL to GPL makes the situation worse in some ways. One has to make sure that The Holy FSF has blessed the license you wish to release your software under is "compatible" since the horrid GPL seems to state (when read cross-eyed and under the influence of a few too many) that a dynamic-linked library is a derivative work.
Naturally, if you managed to Bonoboize Qt, then no-one really gives a shit anymore.:^)
> I now use Mandrake, because of the choice of features it allows
me to install and its wonderful config utilites. HardDrake (hardware detection and config) program rules.
That's what I've been running for quite a while, and, quite frankly, I'm about to switch to something custom or just switch to Slackware. Why? Well, I'm tired of saying, "Why did linuxconf switch that back AGAIN?":^) I'm not sure if it's linuxconf or what, but something keeps changing the permissions on my soundcard, the soundcard's mixer, and the CD-ROM drive. Damn frustrating, that. This could get to be *real* fun when we switch to devfs.:^)
I've been posting the same thing to people for months. WTF is with people? It's not just repeats on Slashdot either. I've seen posts that look like:
"Uh, you're lame; this was on Geeknews like yesterday you fucking lamers"
and the poster thinks they're doing the world a service. In most cases, the post is news to *me.* I'm willing to bet that the post is news to other folks, too. Why do I read Slashdot? Because they post stories I'm interested in and I don't have to go trolling the 'Net. Have I seen this submission before? Maybe...but I don't remember it. Maybe the person is thinking of the nice Linux Journal story on an unrelated wristwatch.:^P
So, yeah, I agree: the "this has been posted here/elsewhere sometime in the last 6 months" post is fucking lame, it's pathetic, and the poster(s) guilty need to die. Now.
Naw, it would only be anticompetitive if they were trying to do something to shut down Jabber, for instance, but this is a case of people writing their own software clients for what is essentially a commercial service and you're essentially connecting to their server.
Think of it this way: you can't just go to a pay phone with a tone dialer and expect to be greeted with a smile by the phone company and the authorities. Granted, the phone company is losing money and AOL isn't, but it's the best analogy I could come up with.:^)
if you are planning to use the Sorensen codec.DLLs to view Sorensen Quicktime files, don't release your software under the GPL. You'll get it from both ends!
A company in the southern part of Illinois (an hour's drive from where I live) is using water towers in a town...they couldn't get permission to build towers (eyesore) but they *could* get permission to put antennae on water towers.:^) Apparently very nice data rates...the only problem is that the southern part of Illinois is a bit hilly, unlike the vast wasteland of central and northern Illinois.
As in, "It hasn't happened yet, but may someday, but until then, it's still not dual-licensed and still only MPL." You'd know that if you bothered to think.
>As someone mention in another post, it's beeing
>*embedded* as a bonobo component.
This is one of those things that's been debated a lot: is CORBA a link, or is it not? Clearly, you need the MPLed Mozilla code to display the HTML in this case, but it's not linking in a traditional sense. This is one of those things that'll be covered (I hope) in a future revision of the GPL (but isn't yet.)
Re:ESR to Linus - I rule!
on
Lawsuits Suck
·
· Score: 1
Yeah, Eric has some good points. However, the reverse can sometimes be true. I just got done working on the production end of a newsaper, and, to tell the truth, sometimes *I* was telling the folks with years of experience what to do...not out of arrogance, but just because I knew what to do. They taught me a lot, but I taught them a thing or too, also.
6. c. If the items are not available to the general public, and the initial developer of the Software requests a copy of the items,
then you must supply one.
...which is an "additional restriction".
*/
OK, so if for some reason the source isn't available to the general public, and if for some reason the original author wants a copy, you have to supply it.
I can only think of a couple of reasons why it *wouldn't* be available: One is if this was part of an in-house project, in which case you wouldn't have to give the code to the entire world. Another reason is if you chose to use the BSD license for your software (I think that's OK...isn't it?). Any other reason and you'd be in violation of the QPL (IIRC.)
So...the GPL requires the source to be open on a public project. The QPL requires the source to be open on a public project. There's no disagreement, as I see it...it seems to me that the QPL wishes to put further regulations on those who choose to use a license other than the GPL, and this should have never been used as an argument against the folks at KDE.
As far as the jurisdiction is concerned, I have to agree with what others have voiced here, which is that I feel that there should be some control over where licensing agreement disputes are handled...I don't want to be dragged to Khazakstan(sp) if I release some GPLed code and some company there chooses to steal my code. I'd like some say, especially since their country might permit such usage and claim the GPL has no power there.
Slashdot Mirror
And, quite frankly, our big public stink will make Sun either settle or pull the product...
...and the loophole will still be there. And it won't get fixed.
Why? Well, my God, fixing the GPL could weaken its position!
No, no, no.
:^)
You didn't read what you posted carefully enough. Go back, read it again. It simply states you *must* cause any work derive from the Program or any part thereof to be licensed under the License(GPL). That's why Sun's committed a violation. They need to either do what you suggest they've already done, or stop their distribution.
I used to see this bit used as an argument against (the formerly non-GPL) Qt. That was a good argument, too, except that it's wrong. It only pertains to a program released under this License or to code derived from the Program already under this License. True, the argument was also that the QPL wasn't GPL compatible, but, c'mon, Troll Tech had given their *permission* in their FAQ.
Of course, this is a dead issue even though RMS won't let it drop.
What about the fact that the drivers are a derived work of the kernel? The COPYING file that comes with the kernel states that using the headers in a user-level program aren't a derived work, true, but surely there's a problem with Bruce's argument.
Surely some sort of code within the kernel has to be used in writing the driver. I cannot say with authority since I have never written a kernel driver. But I'm hoping that this post will spur someone to respond to this issue.
Not for the Hitler page (that's kinda funny) but for the Deez Nuts Tribes page.
:^)
Yes.
Yes.
Yes.
Shazbot!
Dammit!
Hi.
No.
Going offense.
Yes.
Woo-hoo!
I don't know.
Yoo-hoo!
Ah crap!
Defending our base.
No.
Attack!
Ya gotta love Tribes.
I've talked to some marketing folks who still don't understand wtf is wrong with sending direct mail via email.
Shit, that beats the hell out of the new kidneys I bought, and the brain. The new brain works almost as good as the old one, but I'm having some troble with my logic/reasoning. Oh well, a couple more years, and cracking open my old logic book, and I'll be right as rain *tic*
How often are you reading? I read fairly often...and I haven't had major problems in a year.
Talk to Microsoft...oh, wait, you're a Microsoft troll, aren't you?
You know, there will be plenty of flames (esp. on Slashdot) about this article, but there's one thing to keep in mind:
:^) Linux is held together with excellent management skills. I've even seen statements from ESR to that effect.
He's absolutely right.
ESR's description of how Linux was developed isn't really accurate. Many folks already knew it (or, like myself, have read it in other sources.
One could state that this is merely semantics, but it's not entirely correct. It's true that people were not part of some corporation like IBM or Compaq; they were mostly just people who wanted/needed a feature/fix for the kernel, and did it. But that kernel growth/improvement was *managed*, not just some big cloud of code that condensed on a server somewhere.
Absolutely true (and I'm guilty of not having read the article yet.)
Simply stating that it's not true that a project isn't self-organizing is true...and it's false. If there were no interest in the project, and there as a fair amount of dissent between the project coordinator and those who wanted to contribute, the project would die. In a sense, it really *is* self-organizing. IMHO people like Linux are mostly like a managing editor is for a magazine or newspaper--some guiding vision with lots of damage control.
It's not a thing like a traditional company--in a traditional company, people are working towards a common goal so they can get a paycheck. In an open-source project, people are working (hopefully) towards a common goal so they'll have something they'll be able to use.
Really??? That's SWEEEEEEEEEEEEEEEET!!!! :^) This would allow the sort of thing I've been concerned with...at least, I think it will. :^) I'll have to do a re-read of the QPL, but I was wondering if it allowed greater freedom of what you can and cannot use as a license for your "derived work." :^)
Read again--I made no such claim. It was *I* who wished that Troll Tech would release under the LGPL rather than the GPL. I never said that Troll Tech wanted or wished that they could release Qt under the LGPL. By releasing under the GPL rather than the LGPL, they've just made things more difficult for folks who might choose to release software based on Free QT--LGPL makes some of those licensing difficulties magically melt away.
Perhaps the Troll Tech crew is hoping to derive code out of Qt later for use in a non-library. The LGPL won't allow that (check out the LGPL--derived works *must* be a library AFAIK). If that's the case then, well, OK.
>it is always seemed to me that the GNU Public >License exhibits an intolerance to even the >slightest differences in philosophy.
:^) Am I in trouble here? Maybe. But's a pretty ambiguous situation.
Absolutely. There is no tolerance for difference of opinion in the GPL. And if you do a careful reading of the GPL, you'll learn just how *ambiguous* the GPL is, even though it seems to be highly specific. I think the best example is "derivative work."
My reasoning? Yeah, it seems pretty clear that the FSF had a pretty good claim. But the GPL takes the beauty of linked libraries and just stomps all over it. The whole point of a library is that you can have this pre-compiled piece of code and have it work with what you're compiling right now. Did you have to have the source to compile your code? Aside from headers, no. Where's the wording about headers? Er, um...huh. How about dynamic linking? Is the object code of your program compiled with offending object code? Yes and no.
Ooh, here's a case that'll leave you wondering. Suppose I release what is, for all appearances, a source- and binary- compatible GTK clone. Then I release it under a license that doesn't require me to release source, and I distribute it as binary-only. Users install it, because I've made a much lighter, much faster GTK. They run Gnome apps on this. Who's the violator? Me? No, I hired someone to write a specification for me, and based my code on the spec, not GTK source. The GNOME team? Nope...they want nothing to do with me. The user? Well, wait; mine's binary-compatible with GTK. All the user had to do was drop in my libs to an already working system. Where's the violation? There's one somewhere; where is it?
Even better: let's say some company purchases a commercial license for Qt and releases a dynamically-linked version of their closed-source commercial app. I install the app on my system and run it; after all, I installed the Qt RPMs.
>1) You just imagine a hypothetical situation,
:^)
> draw your conclusions about Mr. Stallman's
> reaction and then want us to blame him about
> this reaction YOU imagined.
I believe the author was referring to the QPL/GPL debate, and wrapping it in a thinly-veiled hypothetical situation.
AFAIK the problem pertained to the fact that the QPL was subject to the laws of Norway, and that the GPL doesn't allow that. That wouldn't have been a problem if someone hadn't gotten half-witted and decided that dynamically-linked libraries were derivative works. This means, of course, that I cannot write software for free operating systems, as I'm under the constant fear that, not wanting to release software under the GPL, that any/all authors who have released libs under the LGPL will switch to the GPL (which the FSF advocates.)
Oh well, I've wanted to learn about MFC for a while now.
Exactly.
:^)
But it was a misinterpretation of this statement that caused certain Debian members to go off half-cocked and state that KDE was illegal.
If anyone should be begging for forgiveness, it's the debian team (or at least the offending(offensive?)) members. Now Troll Tech has been forced into releasing Qt under the GPL (NOT the LGPL, as they should have) due to the constant whining (oh, you don't think they were forced? I think they're just trying to save face) and we now have less choice in what license a KDE app can be released under. Want to release something that meets Troll Tech's "free" requirements, yet keep it open in case you want to "go commercial" later? Well, TOO FRICKIN' BAD!
PS Please prove the last statement wrong...it would make me feel better about the world.
is that Qt switching from QPL to GPL makes the situation worse in some ways. One has to make sure that The Holy FSF has blessed the license you wish to release your software under is "compatible" since the horrid GPL seems to state (when read cross-eyed and under the influence of a few too many) that a dynamic-linked library is a derivative work. Naturally, if you managed to Bonoboize Qt, then no-one really gives a shit anymore. :^)
> I now use Mandrake, because of the choice of features it allows
:^) I'm not sure if it's linuxconf or what, but something keeps changing the permissions on my soundcard, the soundcard's mixer, and the CD-ROM drive. Damn frustrating, that. This could get to be *real* fun when we switch to devfs. :^)
me to install and its wonderful config utilites. HardDrake (hardware detection and config) program rules.
That's what I've been running for quite a while, and, quite frankly, I'm about to switch to something custom or just switch to Slackware. Why? Well, I'm tired of saying, "Why did linuxconf switch that back AGAIN?"
Oh, come on, -cycle has got to be the lamest option ever. :^)
I've been posting the same thing to people for months. WTF is with people? It's not just repeats on Slashdot either. I've seen posts that look like:
:^P
"Uh, you're lame; this was on Geeknews like yesterday you fucking lamers"
and the poster thinks they're doing the world a service. In most cases, the post is news to *me.* I'm willing to bet that the post is news to other folks, too. Why do I read Slashdot? Because they post stories I'm interested in and I don't have to go trolling the 'Net. Have I seen this submission before? Maybe...but I don't remember it. Maybe the person is thinking of the nice Linux Journal story on an unrelated wristwatch.
So, yeah, I agree: the "this has been posted here/elsewhere sometime in the last 6 months" post is fucking lame, it's pathetic, and the poster(s) guilty need to die. Now.
Naw, it would only be anticompetitive if they were trying to do something to shut down Jabber, for instance, but this is a case of people writing their own software clients for what is essentially a commercial service and you're essentially connecting to their server.
:^)
Think of it this way: you can't just go to a pay phone with a tone dialer and expect to be greeted with a smile by the phone company and the authorities. Granted, the phone company is losing money and AOL isn't, but it's the best analogy I could come up with.
if you are planning to use the Sorensen codec .DLLs to view Sorensen Quicktime files, don't release your software under the GPL. You'll get it from both ends!
A company in the southern part of Illinois (an hour's drive from where I live) is using water towers in a town...they couldn't get permission to build towers (eyesore) but they *could* get permission to put antennae on water towers. :^) Apparently very nice data rates...the only problem is that the southern part of Illinois is a bit hilly, unlike the vast wasteland of central and northern Illinois.
>Mozilla will have a dual license.
As in, "It hasn't happened yet, but may someday, but until then, it's still not dual-licensed and still only MPL." You'd know that if you bothered to think.
>As someone mention in another post, it's beeing
>*embedded* as a bonobo component.
This is one of those things that's been debated a lot: is CORBA a link, or is it not? Clearly, you need the MPLed Mozilla code to display the HTML in this case, but it's not linking in a traditional sense. This is one of those things that'll be covered (I hope) in a future revision of the GPL (but isn't yet.)
Yeah, Eric has some good points. However, the reverse can sometimes be true. I just got done working on the production end of a newsaper, and, to tell the truth, sometimes *I* was telling the folks with years of experience what to do...not out of arrogance, but just because I knew what to do. They taught me a lot, but I taught them a thing or too, also.
Hasn't happened yet, has it?
Erm,
...which is an "additional restriction".
/*
The main sticking point is/was here:
6. c. If the items are not available to the general public, and the initial developer of the Software requests a copy of the items,
then you must supply one.
*/
OK, so if for some reason the source isn't available to the general public, and if for some reason the original author wants a copy, you have to supply it.
I can only think of a couple of reasons why it *wouldn't* be available: One is if this was part of an in-house project, in which case you wouldn't have to give the code to the entire world. Another reason is if you chose to use the BSD license for your software (I think that's OK...isn't it?). Any other reason and you'd be in violation of the QPL (IIRC.)
So...the GPL requires the source to be open on a public project. The QPL requires the source to be open on a public project. There's no disagreement, as I see it...it seems to me that the QPL wishes to put further regulations on those who choose to use a license other than the GPL, and this should have never been used as an argument against the folks at KDE.
As far as the jurisdiction is concerned, I have to agree with what others have voiced here, which is that I feel that there should be some control over where licensing agreement disputes are handled...I don't want to be dragged to Khazakstan(sp) if I release some GPLed code and some company there chooses to steal my code. I'd like some say, especially since their country might permit such usage and claim the GPL has no power there.
The GPL and LGPL are broken. Let's fix them.