Submitted by “It’s always an interesting day when you get to write a kernel patch, at the urging of Andrew Morton, that notifies the world that non-GPL Linux kernel modules will not work after January 2008 and write some poetry all in the same message.”
More here. Hopefully, many closed-source drivers will be opened during the next year if this patch goes through. Update: Linus responds.
Non-GPL Linux Kernel Modules Banned Starting January 2008?
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSAlert.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
I have sincere doubts that nvidia will open their drivers, but time will tell.
I really don’t need this. I’m voting against this with my wallet. I’ve already purchased an iMac (17″) for one of my daughters for Christmas. I didn’t want to get a Windows machine because I don’t trust Windows anymore. And I refused to get a machine and install Linux, any Linux distribution, because it’s so easy to break and there’s little in the way of quality applications that match what my daughter is used to under Windows and Mac OSX from high school. So I bought a Mac. And as time goes along, I’ll buy more of them for personal use as well as for work.
With the exception of the hard-core user, Linux is going to make themselves unwelcome with these antics. Who wants to get pulled into this zealot’s fight? All most folks want to do is run the hardware they purchased and get the best experience possible with as little hassle as possible. Blocking binary-only “non-free” drivers is the ultimate in hassle. They just want it to work.
And as an embedded developer, I can tell you this: if this really goes into the kernel, I’ll look long and hard at another alternative to embedded Linux, starting tomorrow.
This is not a zealot fight. Kernel developers believe that non-GPL modules are a violation of the license of their code. This is a very simple way they found to help enforce their license.
If your lawyers tell you the Linux developers are wrong with their read of the GPL you can take 2 minutes to modify your copy and use that. You can even distribute it as “Wbeebex” if you want to.
It’s not so simple on other cases, though. Apple goes through great lengths to assure their EULA is respected. Even if the law tells you their EULA is wrong there’s nothing you can do about it, like installing your copy of Mac OS X on a computer not manufactured by Apple.
Microsoft also spends a lot of time to engineer their WGA scheme, which not surprisingly doesn’t work too well. But its purpose is to defend the license of their product.
Some people have different standards. It’s okay for Apple and Microsoft to defend their copyrighted work, but when Linux developers do a very minor thing with a similar purpose then it’s blind zealotry.
And by the way, unlike someone mentioned there’s no “time bomb” on the kernel. Non-GPL modules will continue to work fine after Jan 2008, they will only be blocked on versions of the kernel released after that date.
To sum up:
1) Kernel developers are enforcing the license of their work;
2) People that don’t agree with their read of the license are still allowed to disagree with the kernel developers;
3) People are being given over one year of notice before this change happens.
This is not a zealot fight. Kernel developers believe that non-GPL modules are a violation of the license of their code. This is a very simple way they found to help enforce their license.
This is a zealot fight. Some kernel developers believe non-GPL modules are a violation of the GPL. Others – particularly Linus – believe that you cannot make a blanket statement about that.
What’s not in dispute:
Your company designs a new network card. You decide to code a Linux driver for it. You do this by taking the code for an existing driver in the kernel and modifying it to work with your card. In this case, your driver is very clearly a derived work of the kernel and must fall under the GPL.
What is in dispute:
Same network card scenario as above. This time, your company decides to implement a Linux driver by creating a little bridge code to interface between the Linux kernel and your existing Windows codebase. You release the bridge code under the GPL, but release the rest of the driver only in binary form. Some developers – the ones arguing for this change – insist your entire driver codebase is a derived work of the kernel, and therefore must fall under the GPL. Others – including Linus – feel that your Linux driver is a derived work of the Windows driver, therefore, it falls under the same restrictions as your Windows codebase. Under this viewpoint, only the Linux specific glue code is a derived work of the kernel and must be GPLed.
While I’m all for enforcing the license/copyright – this is bull.
Seriously. Linux and the GPL are supposed to be all about Freedom. Sure, it needs to remain open – and the GPL (though it has a few flaws) does this well (imho).
The problem is that in many cases these are just individuals or small organizations that use these drivers. I doubt many enterprise Linux customers are installing Nvidia drivers (sure, sure, some might).
So…how’s this hurting Linux? How is me installing a driver to make my system as functional (3D wise) as my Windows machine somehow violating copyright?
If Joe Linux installing the NV/ATI driver just to try to make his system comparable is somehow illegal/against the developer’s wishes, then when this happens Linux will lose a lot of users. Will it die? Not likely. I’m afraid it will be relegated to a RISC type place in the OS world though.
Maybe I should start looking into BSD
Then where do you draw the line. Linus intentionally leaves a grey area for practical(sic) reasons. The main reason its been adopted, grown out of all recognition is the Licence.
Do binary drivers hurt linux…absolutely. 3D on Linux Open-source or not is a poor on Linux. Would the situation be better in a years time if the binary drivers were removed?
Edited 2006-12-14 14:15
I think one of the problems here is people are looking for a tidy, black and white solution. There isn’t one.
Most companies don’t find writing open source drivers prudent. Period. Unless they have motivation to do so (gains in marketshare etc), they will not.
ATI/NVidia never struck me as more than token supporters of Linux as it is. Tell them they need to open their drivers, and we may lose what little support we (users of Linux) have.
No matter what, I, as a user, should have the right to control what is or is not on my system. Not some maintainer/copyright holder. Dictate what I can/cannot do with my computer and you’re stumbling into proprietary (eek, drm/trusted computing?!) territory. And defeating one of the chief strengths of foss software!
ATI/NVidia never struck me as more than token supporters of Linux as it is. Tell them they need to open their drivers, and we may lose what little support we (users of Linux) have.
I’m wondering, though…has anyone actually made any serious attempts to convince them to go open-source? I agree that threats won’t work (and may have the opposite effect), but how about trying to reason with them? After all, making their drivers open-source would make things a lot easier for them, as they no longer would have to maintain them.
I wonder how serious the community has been in trying to establish an ongoing dialog with these two companies…
“ome people have different standards. It’s okay for Apple and Microsoft to defend their copyrighted work, but when Linux developers do a very minor thing with a similar purpose then it’s blind zealotry. ”
Actually, Linux fans badmouth MS’s and Apples liscenses all the time, much louder than this. Just look at the uproar over running Vista Home in a VM
That is not enforcement, that is punishment.
Read a bit further down in the thread. Linus himself wrote in explaining how stupid he thinks this idea is, which I would think effectively kills it.
With the exception of the hard-core user, Linux is going to make themselves unwelcome with these antics. Who wants to get pulled into this zealot’s fight? All most folks want to do is run the hardware they purchased and get the best experience possible with as little hassle as possible. Blocking binary-only “non-free” drivers is the ultimate in hassle. They just want it to work.
Talk about short-sighted. You’re failing to see the other side to this issue. The reason your Mac works so nicely is that Apple has the specs and the sources to all the major hardware and all the drivers. The reason Apple and Microsoft can go mucking around to their hearts pleasure with new driver models and new 3D models is because they have the source to pretty much everything. The Linux developers don’t, and that’s a major hinderance. So its not just a zealot issue here, but a very pragmatic one: it’s hard to maintain code that you don’t have the source for, and if you want a first-class experience on your platform, you need to maintain your own drivers.
What I really don’t get is the way Linux and GPL advocates constantly bitch about Microsoft and Apple having the advantage of having all the hardware specs at hand and yet keep on insisting that the quality of their own open source drivers top that of the proprietary ones from Microsoft, Apple and the other 3rd party vendors like e.g. NVidia. If the NVidia driver is this bad and insecure, them why even care about it in the first place? Is it simply because it is ‘unfair’ that some functions are not available in the open source ‘nv’ driver compared to the ‘nvidia’ driver? after all, I bet it’s the 3D accelleration stuff that’s important to both end users and coders here….
Of course the complete specs of a product are necessary for an open source driver to be able to completely replace a closed source alternative at any given time, as long as the functionality set of the open source driver is otherwise inadequate.
At the same time, though, that does not reflect negatively on e.g. Apple and Microsoft. The true problem is that the GPL puts some restrictions on the way proprietary and non-proprietary software come together and because of this, GPL advocates tend to believe that everyone everywhere should obey similar openness policies as they themselves enforce, simply for the same developers to be able to honor the license they themselves have chosen….
This effectively means that e.g. NVidia should open source their drivers under the GPL, in order for kernel developers not finding their copyrights violated, rather than accepting the current ‘driver stub’ approach as being close enough to satisfy their own goals for a project they themselves chose the license for… A project they really want to ‘take over the world’, in turn…
I just don’t see how this can be NVidia’s problem to solve, since it makes it okay for open source developers to loot on corporate intellectual property as this is being open souced by various parties, like e.g. Sun Microsystems, however simply benefitting from the same intellectual property is just not ‘good enough’!
To me it seems like some open source advocates tend to enforce an GPL dictatorship on the world, because since they ‘know’ what is right in this world, everyone should follow. Examples of this way of thinking:
-Broadcom should open up its specs for their wireless chipsets
-Sun should go GPL with Solaris as with Java
-NVidia and ATi should go completely GPL
-Apple should completely open source Mac OS X so open source developers could get their hands on Aqua, Quartz, Quartz Extreme, Quicktime and every other little interesting tidbit under the hood
etc.
But face it: There’s always three sides to every story, yours mine and the truth… And because of the truth existing somewhere in between you and me, co-existance and cooperation ought to be the answer…
There’s no need for everything and everyone in the world of software development to following the same rules…
And concerning the ability to change driver models, I’d like to remind you that this has happened on Linux as well, breaking driver compatibility from time to time… As a counter example, Solaris has had a stable driver interface since at least 1998 allowing you to load drivers from at least that time on a modern Solaris 10 or Open Solaris.
Most of MS’s drivers come from the hardware manufacturer’s, MS just certifies them, and distributes them from winupdate
Let me rephrase this:
“Shut up and stop having opinions. Your semi-public mailing list antics are scaring high school students off your product.”
Unfortunately for you, Linux fails to be a product and few people who don’t care at all about these issues read these lists (until some “news” site posts links to them). They’re of interest to us, but until policy is inacted no one else cares. Even then few people will care…
Anyway though, now that Linus has ended the debate maybe you can go ahead and take your impatient foot out of your impatient mouth and quit getting mad that people who develop for Linux care about the rights of their creative works.
I’m not really sure if this would effect nvidia much or not. Technically, the part of the module that interfaces with the kernel IS opensource. It sort of works as an abstraction layer between the kernel and the actual driver.
Actually the source code that is distributed with the driver set up package is proprietary to NVIDIA and according to the license at the beginning of the file, you’re not supposed to even look at it. It’s sad, because the foundations of an ABI layer would be a good contribution.
What if that layer was LGPLed? Doesn’t the GPL have exceptions for object code that doesn’t interact with the GPL software? If modules only used the LGPLed interface from the module, that might be acceptable (IANAL… thank God).
We have an in-house (at the moment) linux distro that uses closed kernel modules. Instead of making us hope for them to be opened, this, if true, is just going to get me some props on ensuring that the vast majority of our work remains bsd compatible.
… While on the subject:
Just push all the OS dependent code into a wrapper (That you can double license: GPL + proprietary) and hook the proprietary code on top of it. (IANAL, but the double license should shield you against derived work – as long as your taking the conservative view of the GPL)
Yes, it does make life a bit harder (you’ll have to implement your own trees/lists/etc and refrain from using platform/arch only features such as tasklets/workqueues/etc), and adds (a bit) over-head, but it should make platform jumping much easier if all hell break lose.
– Gilboa
It doesn’t matter if nvidia opens their drivers or not, there is a decent amount of work going into writing open source 3D accelerated drivers for nvidia cards:
http://nouveau.freedesktop.org/wiki/
Also, the OpenBSD atheros wireless chipset driver has been determined to be legally acceptable. This means it is in the process of being ported to Linux:
http://lwn.net/Articles/209472/
Edit: Corrected a link
Edited 2006-12-14 03:16
It doesn’t matter if nvidia opens their drivers or not, there is a decent amount of work going into writing open source 3D accelerated drivers for nvidia cards:
You’re right, it won’t matter…because people like me won’t use Linux anymore. Those Open Source drivers have *no* chance of getting anywhere near the performance and functionality level that nVidia provides today on their own.
and how do you determine that? what makes you an expert? why should graphics be so insanely different?
just about all drivers linux have are better than the closed ones on winblows, in terms of stability and speed, why should opensource graphics drivers, when there really are an effort on them be any different? are you saying nvidia/ati are magically having ALL people that knows anything about creating fast graphics drivers? are you really so deluted?
just about all drivers linux have are better than the closed ones on winblows, in terms of stability and speed
Uh, no.
uh, yes.
Uh why?
Do you just believe, or do you know something I don’t?
i have had to deal with lots of hardware during the past years, and the proprietary drivers running on winblows works a whole lot worse than those you find in the linux tree.
I’m sure it doesn’t hurt that they have direct access to the hardware specifications for the devices…
Additionally, writing a driver for an advanced hardware accelerated graphics card, with 3D support, is much more of an undertaking than trying to support a network interface card, for example.
In this instance, “no chance” may be too strong or blanket of a phrase to use; perhaps “extremely unlikely” would be better-suited?
Edited 2006-12-14 05:57
In a year or two, graphics cards will become stream (co)processors, which have a simple CPU-like instruction set (except for vector-valued registers) and can be programmed accordingly. And there’s a good chance that this functionality will be on the same chip as the CPU, with graphics functionality implemented in software on top.
Given that, I don’t see why Linux drivers would be hindered. Surely AMD and Intel are not going to add new x86 instructions which require closed-source drivers to work? (I’m not even sure how that would be possible, honestly.) And if nVidia chooses to resist — which is unlikely, given that their new chip works exactly like this — then the market will leave them behind, since people will already have processors which can do all the graphics computing they need.
> Surely AMD and Intel are not going to add new x86
> instructions which require closed-source drivers to
> work? (I’m not even sure how that would be possible,
> honestly.)
While I agree that this is unlikely, it’s not impossible, after all that’s what Transmeta did with their VLIW: the VLIW ISA was ‘closed’ and there was a translator from x86 to the VLIW ISA.
There could be a similar thing from GPU: your program would use a ‘public binary interface’ which would be translated to the real ISA underneath, after all that’s what CPU with microcode do, whether the translatior is microcoded inside the CPU or an interpretor software is just an implementation detail.
GPU instructions wouldn’t be very useful unless you could use them directly. SSE 1/2/3 are useful precisely because they give you instructions for manipulating vectors directly. So the ISA would have to be modified in some way.
Now, if the public interface was translated to some other instruction set, then the other instruction set would have to exist at the microcode level to be practical. Even RMS doesn’t object to microcode. If it isn’t physically possible to reprogram a chip, then the GPL’s freedoms are not necessary or applicable.
just about all drivers linux have are better than the closed ones on winblows, in terms of stability and speed
I don’t know about the other ones, but for example the nv driver doesn’t beat nvidia’s one. nvidia’s one atleast supports all the card’s functionality. And the same applies to the open-source r200 driver: my Radeon 9100 doesn’t get tv-out or pixel shaders when using the open-source ones, and not having pixel shaders sucks. I know the recent ati-drivers don’t support my card anymore, but still, when they did, they did a better job at it than the open-source one.
I know the recent ati-drivers don’t support my card anymore, but still, when they did, they did a better job at it than the open-source one.
That’s exactly the problem of closed source drivers. When you buy a video card, you can use it as long as you want, with any operating system and any version. I have a GeForce 2 in this PC. Some would argue this is a very old system (I bought it on December 2000). Still, I’m running the latest Debian Sid on this system.
I’m pretty sure, sooner or later, NVIDIA will stop supporting this card (for the moment they provide nvidia legacy drivers, but for how long ?). There’s nothing really wrong with that. It would cost them too much money to write a driver for such an old hardware almost nobody still uses nowadays. That’s what’s good with opensource drivers. I’m pretty sure I’ll be able to use my geforce 2 with full 3D acceleration in the future. I wish nouveau success
.
“are you saying nvidia/ati are magically having ALL people that knows anything about creating fast graphics drivers? are you really so deluted?”
In fact… they do. It’s THEIR hardware, after all, so THEY have the people who know it’s design and how to code a driver for it in order to get the most of it.
Todays graphic chips are extremely complicated pieces of hardware so it’s a little bit different than making a driver for some network card or something like that i guess…
im sure the people who designed all sorts of other hardware are also the experts on it.. just too bad that those companies drivers arent the best?
i got news for you, nvidia/ati doesent have monopoly on having smart and skilled people
No, they certainly don’t.
The problem of course is just how complex these cards are nowadays.
Not to mention, I’m not sure what open source developers have as a reference point.
They’re writing from scratch and reverse engineering. Guess what? This is HARD. I don’t care how brilliant you are – grab a brand new Geforce and start writing a driver – it’s going to take a LOOONNNGGGG time to even make something that functions well. Something that’s stable, fast, and supports EVERYTHING in the card?
I have a lot of faith in smart people. Unfortunately, smart people also have lives, jobs, etc. Plus a lot of people have other hobbies or projects they contribute to.
Don’t worry, you’ll still be able to use Linux with proprietary drivers. The copyright holders can’t tell you what you can do with their work as long as you don’t redistribute it.
Thats cool. So I can just remove any checks in the code and redistribute the kernel. My friend can then release his closed drivers ok
Anyone here think that we’ll move from a model of
yum install kmod-nvidia
to
yum install kernel-nvidia
This is not enforcement as some people suggested but obfuscation and irritation inserted into complex code to cause pain to people who use the code in manner unpalletable to the people who once gave it away … whilst doing nothing to legally or practically prevent such usage. its very silly
<quote> It doesn’t matter if nvidia opens their drivers or not, there is a decent amount of work going into writing open source 3D accelerated drivers for nvidia cards:
http://nouveau.freedesktop.org/wiki/ </quote>
From the nouveau site you linked to: “Currently, nothing works.” They might be putting some work into it, but there is little evidence of a usable driver any time soon. I very much suspect that it does matter, as it will take them years to produce a functional driver, and it is likely to be very out of date and lacking in functionality compared to the nVidia one.
I would say that if Linux moves to this kind of restrictive, draconian policy of controlling what people use their computers and software for, people will move away from Linux as fast as you can say BSD, Vista or Leopard. Restricting people’s freedom is not going to win Linux any new friends.
Linus is on the ball here, it is a shame about the other kernel devs.
PC-BSD and FreeBSD are very rapidly approaching the point of becoming a viable desktop OS, and are not encumbered by lots of silly licensing restrictions. If Linux gets carried away with some silly idealistic notion of “freedom” that is ironically the exact opposite of what freedom means to most people, Linux will fade back into being a server only OS, and has no chance of making it on the home or office desktop.
Just a thought:
The GPL expressly permits me to use the software for any purpose whatsoever. Indeed, this is enshrined as the FSF’s “Zeroth Freedom”. So presumably this change can’t stop non-GPL kernel modules from being loaded *full stop*, as that would be a violation of my GPL rights to f–k about with the program as I choose.
I can see it would have an effect on the distribution of kernels with (for example) the Nvidia module built in, but how does it stop the present use-case of people just installing 3D/wireless drivers themselves?
There is nothing that would prevent you from removing the licence check routine from the kernel, nor from distributing such modified version… The GPL allows such modifications, after all. If there is a demand, it will happen.
Of course, it would be quite hard to get the cooperation from the kernel developers for analysing a crash using such kernel, but human cooperation isn’t a requirement of the GPL.
prevent you from removing the licence check routine
What keeps confusing me is M. Shuttleworth’s announcement
that Ubuntu will have binary drivers in kernel. How is that possible ? GPL prohibits that, because there is no source code available for them, right ?
DG
What keeps confusing me is M. Shuttleworth’s announcement that Ubuntu will have binary drivers in kernel. How is that possible ?
What will happen is that the kernel and closed-source modules will be distributed separately on the same CD-ROM, and will only be combined during installation.
What keeps confusing me is M. Shuttleworth’s announcement
that Ubuntu will have binary drivers in kernel. How is that possible ? GPL prohibits that, because there is no source code available for them, right ?
AFAIK, it is NOT. You can legally distribute them seperately even if they are on the same delivery medium (i.e. the same CD, the same ftp site, etc) then install them on the users machine.
BUT, you cannot distribute the kernel built with the modules, i.e. compiled-in/statically linked. The kernel modules are just object code and NOT linked until loaded (insmod).
Anyway, the GPL only applies when you try to distribute the software.
Distribute, copy or modify. If you modify Linux by loading a kernel module you have to comply with the terms of the GPL.
//Distribute, copy or modify. If you modify Linux by loading a kernel module you have to comply with the terms of the GPL.//
No. AFIAK you can modify your own copy of GPL software to your own heart’s content, and never have to publish your modifications if you don’t feel inclined to do so. The GPL only applies if you try to redistribute the code.
//Distribute, copy or modify. If you modify Linux by loading a kernel module you have to comply with the terms of the GPL.//
No. AFIAK you can modify your own copy of GPL software to your own heart’s content, and never have to publish your modifications if you don’t feel inclined to do so. The GPL only applies if you try to redistribute the code.
Precisely. The irony is that the GPL is basically designed to prevent software developers from doing precisely what the kernel devs are trying to do: deny users the ability to modify and execute the code as they see fit.
thats a lie, the gpl is there to prevent companies like nvidia from releasing illegal nonfree modules for a gpl’ed thing.
Uh, no. As a copyright license, the GPL is concerned with one thing: redistribution.
It’s not illegal to release a closed-source driver for the Linux kernel. It’s not illegal to link said driver to the kernel.
What *is* illegal is to redistribute the kernel and driver once they’re combined.
To be more precise, what is claimed to be illegal is redistributing the linked kernel + driver.
This is Greg KH’s position. He’s wrong, as Linus has pointed out several times.
This is one of the few areas of the GPL in which there is related case law, and what case law there is rules against Greg’s position and for Linus’
Yes, you’re totally right. It’s not even clear if redistributing the linked kernel+driver is in fact illegal. I wasn’t aware that there were legal precendents against it, though…which ones were you thinking about?
In any case, this seems to have been much ado about nothing, but at least it gave the OSAlert anti-Linux crowd something to get excited about…
I don’t have the case handy, but as I said it’s indirectly related. There is a bit of case law on the definition of ‘derivitive work’ with respect to software and it makes it clear that Linus’ position is closer to that taken by the courts than Greg’s.
Basically, a work isn’t “derivitive” just because it’s linked with another piece of code, especially if the interaction is clearly delineated by interfaces.
This is where Linus’ “grey area” comes into play. As he has stated, a module that was written entirely for another OS but that is coupled to the Linux kernel via a shim is not going to be seen by the courts as derivitive.
The Nvidia drivers, for example, fit that model fairly well.
Basically, a work isn’t “derivitive” just because it’s linked with another piece of code, especially if the interaction is clearly delineated by interfaces.
In fact, I believe this was used as an argument against SCO’s claim in the IBM/SCO trial. I could be wrong, though, but I seem to remember something about this.
except its not illegal. Unless of course nvidia are distributing someone elses copyright material without license. Their own private *use* of the kernel does not contaminate their code. Think about it this way
I write code (its mine) lets call it “A”
I compile and insert it into the kernel (it becomes GPL at this point) lets call this /copy/ of my code “B”.
Now A has never touched the kernel and I can do erxactly what I like with it. just because I type #include linux.h doesn’t make it GPL.
I don’t have any obligation for B because I didn’t distribute it.
The **only** ambiguity is a compiled version which used linux.h in the build process. If you belive using an include file makes it a derived work, then you’re skating on very thin ice – and I bet the .h file could be clean room reimplemented anyway.
The distinction between GPL and LGPL is alsoa bit suspect, but LGPL is explicitly unmuddying the water for the benefit of the downstream user.
No, it’s to encourage the sharing of code, for the betterment of the code, at least until GPL 3
“thats a lie, the gpl is there to prevent companies like nvidia from releasing illegal nonfree modules for a gpl’ed thing.”
Whether or not non-gpl modules are illegal is precisely what is up for debate, and Linus says they’re not. If you can’t live with a few non-open-sourced things on your system, I hear that Hurd is looking for some developers.
No. You don’t have to comply with the GPL unless you distribute the GPL’ed software.
Isn’t it true also the fact that the whole time that the code is on that distribution CD, the kernel and the module are not, in fact, linked? And that the only time that the code is linked with the kernel is when it’s run, by the user? Can’t one imply that the users are the only group that this could be enforced on? I don’t think they could get NVIDIA for distributing some proprietary source and a binary blob even if the point of the code is to link with Linux.
Isn’t it true also the fact that the whole time that the code is on that distribution CD, the kernel and the module are not, in fact, linked? And that the only time that the code is linked with the kernel is when it’s run, by the user? Can’t one imply that the users are the only group that this could be enforced on? I don’t think they could get NVIDIA for distributing some proprietary source and a binary blob even if the point of the code is to link with Linux.
The issue and point of contention is that the act of compiling a proprietary driver against the kernel makes it a derived product of the kernel and therefore subject to GPL licensing. I say it’s a point of contention because there’s some debate about what constitutes a derived work in software under copyright law. In some cases it could be clear cut, in other cases not so.
It really becomes a grey area with drivers like nvidia, which are compiled wrappers against a binary blob that is actually designed to work on Windows and *nix interchangeably. In most cases the user simply downloads the source code for the wrapper, compiles it themselves against their own kernel (via the installer script) and then really, the GPL doesn’t come into play because it’s not distributed already linked.
nVidia however, does provide binary modules for download for specific kernel builds/distros. Under the view of devs like GKH, that is clearly illegal. Yet nvidia’s legal department apparently has a different POV.
You also have an issue with projects like ndiswrapper, which compiles directly against the kernel but links to a proprietary Windows driver. Again, many would argue that a proprietary driver designed for Windows couldn’t be considered a derivative work of the linux kernel, yet again some of the devs feel the issue is black and white and that ndiswrapper is not compatible.
Linus is probably among the most pragmatic of the developers in saying that he’s simply not sure where the line is drawn and that it could vary from jurisdiction to jurisdiction based on interpretations of copyright law.
I would tend to agree that vendors distributing completely self-contained proprietary binary modules precompiled against specific kernel versions would be a violation of the GPL. I’m not so sure I’d agree that distributing precompiled GPL/GPL-compatible wrappers or shims linking to an OS agnostic binary blob is necessarily the same situation. And there is certainly no violation in providing the source for the wrapper and allowing the user to compile and install it themselves. The GPL doesn’t apply to source code, only binaries.
Anyways, just my 2c.
Exactly. Are kernel modules statically linked? If they are, then compilation would result in linking. If they are not, then linking occurs at runtime, in which case compiling would not be considered linking.
If you ask me, someone needs to work on a generic LGPL ABI-compatible shim that is unrelated to any non-GPL projects.
Actually it’s quite simple. A developer can still release binary drivers for Linux, because at the end of the day, we as users don’t have to accept the GPL anyway. So even if the binary drivers are in violation of the GPL, there is a clause that says we can reject it, and we are ok as long as we don’t copy it.
IMHO I think if this goes through linux will slowly die. We’re supposed to trying to get the support of hardware companies, not turn them against us. We might as well say GNU/Linux should not run commercial software at all.
One should also note that linux is irrelevant because at the end of the day we run GNU with Linux as the kernel. It can easily be replaced by something else like: Mach, BSD, Hurd… Hell even the NT kernel could do it.
That’s entirely true, and I’ve heard numerous comments about people desiring to move their stuff to BSD. The kernel should’ve been LGPL or GPL with a linking exception from the start, and I think that’s what Linus is thinking right now. There is no reversing it I’m afraid, as Linux has far too many contributors to line up copyright holders with a new license. Honestly, this is just a vendetta for personal gain against non-free software that hurts the community.
What’s worse is that most of the argument is based on emotional arguments! You see kernel devs saying “their doing crazy crap in userspace” which doesn’t affect them. Simply enough, the tainted bit avoids kernel developers needing to deal with problems caused by binary drivers, and who are the victims of this crime? It is certainly not kernel developers like Greg Kroah Hartman who (afais from his site) has contributed a minute amount of code to the kernel. I’d like to see the kernel developers who want to be rid of binary modules put up a website, name themselves and their code, so we can patch it right out. I don’t want to run a kernel with code written by NAZIS in it… sorry devs but you are Nazis.
Linus was right on when he said this was hipocrasy. The Nazis even went so far as to underplay Linus’ contributions to Linux! Come on, if he hadn’t written it, you’d all be working on TPS reports at some crazy office!
I don’t want to run a kernel with code written by NAZIS in it… sorry devs but you are Nazis.
Godwinned!
http://en.wikipedia.org/wiki/Godwin%27s_Law
Oh, that’s priceless, good stuff
It is certainly not kernel developers like Greg Kroah Hartman who (afais from his site) has contributed a minute amount of code to the kernel.
I don’t agree with Greg KH’s zealotry (“BINARY DRIVERS ARE ILLEGAL!”), but in fairness he’s a very competent dev and has contributed some important work to the kernel, particularly the usb and udev subsystems, IIRC.
Deride his politics if you will, I know I do, but don’t demean his work; most people using linux are benefiting from it.
Yes you’re right. It seems his contributions are merely few in number and large in size. Sorry Greg–please mod down that post. The reason this issue gets me riled up is that the same people who fight DRM with all their might, yet we turn around and try to make our code enforce the legality!
I’m not so sure if I understanded well that non-GPL modules will not work with the kernel or that they will simply write something to screen saying they’re not GPLed? Frankly I’m seriously thinking about moving to BSD.
[EDIT]As people look like saying non-free driver will simply not work anymore, I simply think is somehow a misunderstanding or a prank, because Linux will simply alienate itself a lot of hardware builder by doing this, by winning not so much. I mean, my NVidia driver on Windows and Linux are closed and they do the job exactly the way I want. I don’t want them to not work anymore because some zealot pretend it’s better for GPL or open source. I want desktop Linux, I want something working AND supported by the hardware manufacturer. I just hate all kind of fanatism. [/EDIT]
Edited 2006-12-14 03:31
I’m not so sure if I understanded well that non-GPL modules will not work with the kernel or that they will simply write something to screen saying they’re not GPLed? Frankly I’m seriously thinking about moving to BSD.
For a long time, modules have had a symbol in them specifying the license. If the symbol didn’t say GPL, the kernel would still load it, but would print a message warning that you were loading binary code and your kernel was tainted. Once you loaded a binary module, a tainted flag was turned on in the kernel, which would never be cleared. If you sent the dump from a kernel panic to a kernel dev, the first thing they would do was check that flag. If it was set, most kernel devs would pretty much ignore your problem.
I think at this point, they’re just giving a scarier warning message, but that message does give a clear date in the future where the warning message will instead become an error.
I think at this point, they’re just giving a scarier warning message, but that message does give a clear date in the future where the warning message will instead become an error.
With the vanilla kernel source.
What’s to prevent distributions from patching out the exclusion on their own distributed kernels? The GPL itself protects the right to bypass this type of GPL enforcement mechanism.
I don’t want to be screwed by windows and linux only to be screwed by mac. I was really impressed by mac until I realized how much locked in i would be-not just by software but by hardware too
full of sound and fury, signifying nothing.
Everyone has expected such a move for over a year and nobody’s going to change their stance on opening or closing their drivers as a result of it.
yes I want more apps to be open source. but don’t *force* it on those who are simply unready to go all the way. I hate this extremism.
BSD, MIT, MPL-that’s the way to go. And I don’t go this idea that Microsoft will just take the code. They don’t need the code because they’ll just take your ideas and implement them in a way that suits their product. All GPL does it make it harder for bsd/mit/mpl projects to share.
Actually, your criticism of the GPL here is misguided.
Contrary to what some of the kernel developer say, it is not illegal to combine proprietary software with GPLed software…it may be illegal to redistribute the combination, but as others have specified, the GPL allows you to use the software as you see fit (and that includes modifying it, including by linking non-free modules).
Copyright law is all about redistribution. If you don’t redistribute, you can do pretty much everything.
The GPL is a good license, like the others you mentioned. The important thing is to understand what software licenses are, and what they aren’t.
Actually, your criticism of the GPL here is misguided.
Yeah, he broke the cardinal rule of expressing himself about the GPL. Tsk, tsk. Anything which doesn’t conform with “the community’s” shared ideology is “misguided”, according to the zealots…
Instead of attacking me, you could actually have indicated where my reasoning was flawed in your opinion. Since you didn’t, and instead went right to argumentum ad hominem and strawman arguments, then I’ll conclude that you have nothing of value to say, and are simply trolling this thread (as you usually do).
Can’t you just take your antisocial behavior somewhere else?
Actually, you probably deserved to be attacked a little bit. It’s not misguided to disagree with someone. The rest of your argument was pretty correct though regarding the GPL. Remember, “good” is in the eye of the beholder.
Well, I respectfully disagree. In my opinion this particular criticism was misguided in the sense that it was based on an erroneous interpretation of the GPL (i.e. that the GPL somehow restricts what you can do with the code in the privacy of your home).
In any case, if I had made an error or a false statement, the rational response would have been to point it out, not to attack me and present what amount to strawman argument.
“Good” may be in the eye of the beholder, but logic is not subjective. Either an argument is correct based on the available information, or it isn’t. Debating expressely serves to determine the validity of arguments, and logical fallacies such as ad hominem attacks and strawman arguments should always be avoided.
Well, I respectfully disagree. In my opinion this particular criticism was misguided in the sense that it was based on an erroneous interpretation of the GPL (i.e. that the GPL somehow restricts what you can do with the code in the privacy of your home).
His criticism wasn’t misguided unless he didn’t understand that the viral effects of the GPL don’t kick in until distribution time. That is unlikely.
I would wait until more details of this emerge. I would highly doubt that the kernel maintainers would do something like this in light of Microsoft’s recent antics. Barring non-free blobs from the kernel is a BIG mistake and will drive people away from Linux and F/OSS. These guys need to be careful.
In other news, Linux users move away from the &*(@#$&(* Linux platform to one where they can actually *use* the hardware they paid for.
Say, Solaris, BSD, Windows, etc.
Linux, the platform that is about the license, not the users.
Edited 2006-12-14 03:49
If you read the threads, it really has nothing to do with being a GPL fanatic. The developers don’t want to expose IRQs (and other internals) to userspace because that allows for some really stupid things to happen. (Think if an IRQ with your hard drive is shared with an userspace driver. The users space driver does something stupid and all of a sudden your hard drive doesn’t work and your systems crashes) Closed source modules reuqire access to stuff like this and it causes real headaches for the kernel devs.
It’s not so much a matter of licensing issues, so much as it is a matter of the differences betwen how open source and closed source module code interfaces with the kernel. GPL licensing just provides a handy way to put a stop to bad programming practices.
Edited 2006-12-14 04:03
The 2 sections at http://www.kroah.com/log/2006/12/13/#2006_12_13 seem to be closely related….
After seeing the entire discussion (at http://thread.gmane.org/gmane.linux.kernel/475654/focus=475796) it seems like they’re pushing “binary blobs” out of the kernel and into user space, where the damage they can (potentially) do can be minimised.
Linux is becoming more and more like a micro-kernel every day…
If everyone reads the list
A few posts down Linus clearly expresses his disdain for this and the hypocrisy of it
In other words, you guys know my stance. I’ll not fight the combined
opinion of other kernel developers, but I sure as hell won’t be the first
to merge this, and I sure as hell won’t have _my_ tree be the one that
causes this to happen.
So go get it merged in the Ubuntu, (Open)SuSE and RHEL and Fedora trees
first. This is not something where we use my tree as a way to get it to
other trees. This is something where the push had better come from the
other direction.
Because I think it’s stupid. So use somebody else than me to push your
political agendas, please.
Linus
Now I’m not so worried
Edited 2006-12-14 04:50
I somehow doubt that any major distro will merge this fix by itself.
As such, I doubt that I’ll ever make its way into the main-line tree.
– Gilboa
Hmmm, how analogous to a DRM implementation is this.
“We won’t let you run this driver on your system”.
Hmmmmm.
Tp.
It feels like Linux people are closing ranks on themselves. Is this some sort of a move to protect themselves from a Microsoft + Novell world or just another re-tread of the UNIX wars from days gone by.
This all seems so familiar to me.
Like a George Orwell novel in which we all have equal rights and freedoms but some have more than others.
Dennis
Why?
What happens when you become dependant on a bunch of closed source modules you can’t operate without them? Doesn’t that undermine much of what OPENSOURCE is all about? In the kernel, closed source modules are pollution. Get them out now while it’s still easy. For those that run Linux/GPL software because they want the code open do care about this.
There’s nothing stopping the user from installing them afterwards, either. This is only about distribution and this move is meant to protect the user, that means you, from someone inserting something buggy and/or unsecure into your system.
I commend the kernel guys and Mr. Morton for thinking about me, even if it’s not directly.
I understand what you’re saying, and don’t completely disagree. However, how would this work out for non-GPL but still “free” licensed kernel modules? I’m thinking in particular of openafs, which is under the IBM public license (and which we rely on _heavily_ at my work, as to a number of major universities, government institutions, etc.). It’s free, it’s open source, but still the kernel flags it as tainted…
Would this stop running under the above scheme?
This move is not to protect me from myself, but it does stop me from using my hardware to it’s full potential. A lot of closed source companies do not have opensource drivers, and never will, and so I have to a) cripple my linux install to be compliant, or b) jump through more hoops then is needed to get things working, and it wasn’t that easy to begin with. It just means that Linux has one less user, I’ll stick with FreeBSD and Windows from now on, I know my rights won’t change for no good reason there
“This move is not to protect me from myself, but it does stop me from using my hardware to it’s full potential.”
Your right, it’s not to protect you from yourself. It’s to protect you from whom you receive your software from.
You are also wrong it “stops” you from using your hardware to it’s full potential. Most distros use some form of aptget with a repository. It’s a simple matter after installation to install closed source binaries if you wish (and it takes only about an hour to learn how to install software, even from source, I learned and I didn’t have a clue before I started).
This is effectively the same as Windows. The last time I installed I had to install a bunch of drivers I had to go out and get from the vendors because some of the drivers I got off of the Windows update site were dated and incomplete. Installing drivers is not that big of a deal.
The important difference is the DISTRIBUTION of closed source with opensource. The difference is you can verify what you installed on your computer as safe and stable if you can see the code if it’s all open source code. You can’t if there is a ‘black box’ sitting in the middle of the kernel, the kernel no less. If you accept the compromise you seek, then your stuck having to trust your distributer they didn’t put something in your software.
Maybe you need a wakeup call, but the ONLY thing that keeps a distributer honest is the ability of the user, by taking the code that was received, and anaylizing it as safe. You don’t get that with closed source, no matter how much you trust the distributer. Why do you think so many of the people who actually care about open source are against putting stuff in the kernel?
This is another thing to consider, Linus has his opinion, yes, but he as far as I can see, is not an opensource supporter. He didn’t choose the GPL because it was the product of opensource ideals, he chose it because it was the best license at the time he felt that would force the sharing of derivative work. That’s not a bad thing, but it’s easy to see his choice puts him at odds with the FSF because their goals are not the same. But that’s beside the point of why this is bad.
Why this is bad is it’s the death of a thousand cuts. The point of opensource is it, as a user, allows you to verify the software on your computer is safe, and if it’s not, to fix it. And by safe, I mean security and the preservation of your data. If you allow all sorts of closed source black boxes in the kernel, who would motivated to release open source drivers? Right, no one. They won’t support their hardware and the same uphill battle that’s always taken place, the reverse engineering of hardware to support it by the opensource community, will continue.
Linux has become a very decent kernel, few deny that. There are many who would like to make use of that kernel but for various reasons, from proprietary to criminal intent, from patents to trade secrets, from whatever to whatever, all those people don’t give a hoot about you or your data. The only person who does is you. So why argue away your ability to protect yourself?
But everytime you accept the kernel with chunks of it closed off, you accept that you don’t care about your security and the safety of you data. This isn’t about religion or a cult as some would charge, this is about preserving your ability, as the user of the software, to fix it, to make is secure as possible, and not to be held by a third party just because you use their software.
There is a lot to disagree with when it comes to Stallman. But I do share his wisdom in seeing that if you control your software, that means you also control your hardware and more importantly you contol your data. If you don’t control your software, you don’t control your hardware and you can’t gaurantee your data.
Linus’ vision seems to be more let the chips fall as they may because he has confidence humans will shy away from things that make doing stuff difficult naturally, like DRM. And he’s right, in the long term. But in the short term, he’s not the one suffering from lose of control over their software, hardware, and data.
I’m ranting, and I’m sorry, but it does amaze me so many are so willing to argue away their only protection for the sake of convieniece. Especially when it concernes the kernel.
The last time I installed hardware under Windows, I stuck the CD that came with the hardware and installed the drivers, I didn’t have to look anywhere.
What you haven’t seen, is that the hardware supplier has had it’s share of hoeadache to locate and find the drivers and put them n cd.
Now, compare apples with apples. get a linux distribution and a windows OEM. Still as easy or did you have to look everywhere?
We build systems on specs also for windows and the biggest problem machines are the windows ones..
More than likely, that driver is not the lastest driver and you’ll still either have to look on the windows update site or more importantly, the manufacturer’s website for the latest driver.
But that’s not the thrust of my post. I hope that’s not all you got out of it.
I wonder how many of the distros will exercise their GPL granted right to remove that code for the sake of their customers.
Tp.
My own personal estimate? All of them.
It’s not just about distribution.
When implemented in Jan 2008, the patch will fail to load any driver that has not got the gpl var set.
If I run a distro with that patch in it, as of Jan 2008, I will not be able to load a non-gpl driver.
Tp.
“What’s to prevent distributions from patching out the exclusion on their own distributed kernels? The GPL itself protects the right to bypass this type of GPL enforcement mechanism.”
Thank God. You can bet that’s one open source project I’d contribute money to. That’ll be the first patch I apply to my vanilla kernel after Jan 1, 2008.
What a stupid move. Talk about biting your nose to spite your face. Hopefully this won’t force a mass exodus back to Windows just so people can get their hardware to work.
I love Linux because I can do anything I want with it. *Anything*. I don’t care what their reasoning is; if they’re going to do this, then they might as well adopt GPL3 since it will have practically the same effect.
I fail to see how this is a good thing?
i fully understand, with which the intentions it is being done. and understand them, but: wont this stop some venders from dealing with Linux at all? many venders dont do GPL and or opensource, they provide they software via binaries. if that is blocked then how is that a good thing? OR is this an attempt to force everything into userpace?
-Nex6
I blogged about this tonight:
http://eugenia.blogsome.com/2006/12/13/no-closed-source-drivers-on-…
I don’t necessarily agree with the general direction, but at least that would be a clear position on a matter that needed a position.
Edited 2006-12-14 04:34
If Nvidia stops making a driver for Linux, then it will soon stop making a driver for FreeBSD. FreeBSD’s driver is mostly the Linux driver with some hooks to the FreeBSD kernel.
Unless FreeBSD gains some momentum of its own, the market share for 3D accelerated graphics on it is just too small.
Neither of the Unix drivers is made because of market share. If you listen to the audio blog from nvidia you’d know that they basically did it out of curiosity and it worked so well they decided to let it go.
They don’t do linux or freeBSD or solaris drivers because of market share. Most people use ALSO windows anyhow, and MOSTLY windows for their 3d needs (games).
Edited 2006-12-14 08:55
I have tested PC-BSD, and it is realy easy to install and to use. I have removed it in favour of FreeBSD, because I realy wanted to learn something about the system. But the future for PC-BSD is bright. Hardware vendors will not ignore it.
DG
Actually, some FreeBSD developers started out making a driver as a wrapper around the Linux driver. The driver from Nvidia is not a wrapper (*number); it is a native FreeBSD driver (*free man): http://us.download.nvidia.com/freebsd/1.0-9631/README/chapter-04.ht…
* I could not help “The Prisoner” (http://www.imdb.com/title/tt0061287/) reference.
In the past I’ve had hardware that has become ignored by the manufacturer. A year after the 2.6 kernel was released my wireless card manufacturer still only offered 2.4 binary kernel drivers. I sent them emails and never received replies and was forced to buy another wireless card.
With closed source drivers we put our faith that companies will continue to update their drivers. Once a product is no longer supported a company may stop updating the driver. This is not only annoying, but may prove to be dangerous. Say an exploit is found that compromises the security of the kernel. The system then becomes insecure until someone can get the manufacturer to release an updated driver. If the company no longer exists or nobody is working on that driver anymore then the user may be SOL.
With open source drivers users can feel safe knowing someone can continue working on their hardware drivers long after a product is released. New kernel features can be taken advantage of. For example if a soundcard manufacturer believed there was no point on supporting ALSA in their driver then users would be stuck using OSS. Or if NVIDIA decided not to implement texture_from_pixmap users wouldn’t be able to run AIGLX.
Also remember some companies put Linux users second. Look at how long Linux users have had to wait for Adobe Flash 9, and even then it’s still beta software. If Flash was open source maybe we’d have Flash 9 by now.
I believe an all open source kernel would be more secure for the user and would allow for a better experience.
I don’t think anyone is arguing that having open-source drivers isn’t a good thing. Unfortunately, the reality is that binary drivers are here to stay and if you want to use your hardware to its fullest potential they’re usually your only option.
I guess I’ll have to ditch linux to play games and go back to windows.
This patch is a fantastic way to get people to switch to windows.
yes, I was enjoying excellent frame rates using Cedega to play my windows games under linux.
however, it will not make me switch back to linux, instead, it will make me get behind the people writing the open-source driver, and put public pressure on Nvidia to open source “A” driver…. it doesn’t have to be their latest driver… just open source one driver…
From Linus further down the list:
Btw, I really think this is shortsighted.
It will only result in _exactly_ the crap we were just trying to avoid,
namely stupid “shell game” drivers that don’t actually help anything at
all, and move code into user space instead.
What was the point again?
Was the point to alienate people by showing how we’re less about the
technology than about licenses?
Was the point to show that we think we can extend our reach past derived
work boundaries by just saying so?
The silly thing is, the people who tend to push most for this are the
exact SAME people who say that the RIAA etc should not be able to tell
people what to do with the music copyrights that they own, and that the
DMCA is bad because it puts technical limits over the rights expressly
granted by copyright law.
Doesn’t anybody else see that as being hypocritical?
So it’s ok when we do it, but bad when other people do it? Somehow I’m not
surprised, but I still think it’s sad how you guys are showing a marked
two-facedness about this.
The fact is, the reason I don’t think we should force the issue is very
simple: copyright law is simply _better_off_ when you honor the admittedly
gray issue of “derived work”. It’s gray. It’s not black-and-white. But
being gray is _good_. Putting artificial black-and-white technical
counter-measures is actually bad. It’s bad when the RIAA does it, it’s bad
when anybody else does it.
If a module arguably isn’t a derived work, we simply shouldn’t try to say
that its authors have to conform to our worldview.
We should make decisions on TECHNICAL MERIT. And this one is clearly being
pushed on anything but.
I happen to believe that there shouldn’t be technical measures that keep
me from watching my DVD or listening to my music on whatever device I damn
well please. Fair use, man. But it should go the other way too: we should
not try to assert _our_ copyright rules on other peoples code that wasn’t
derived from ours, or assert _our_ technical measures that keep people
from combining things their way.
If people take our code, they’d better behave according to our rules. But
we shouldn’t have to behave according to the RIAA rules just because we
_listen_ to their music. Similarly, nobody should be forced to behave
according to our rules just because they _use_ our system.
There’s a big difference between “copy” and “use”. It’s exatcly the same
issue whether it’s music or code. You can’t re-distribute other peoples
music (becuase it’s _their_ copyright), but they shouldn’t put limits on
how you personally _use_ it (because it’s _your_ life).
Same goes for code. Copyright is about _distribution_, not about use. We
shouldn’t limit how people use the code.
Oh, well. I realize nobody is likely going to listen to me, and everybody
has their opinion set in stone.
That said, I’m going to suggest that you people talk to your COMPANY
LAWYERS on this, and I’m personally not going to merge that particular
code unless you can convince the people you work for to merge it first.
In other words, you guys know my stance. I’ll not fight the combined
opinion of other kernel developers, but I sure as hell won’t be the first
to merge this, and I sure as hell won’t have _my_ tree be the one that
causes this to happen.
So go get it merged in the Ubuntu, (Open)SuSE and RHEL and Fedora trees
first. This is not something where we use my tree as a way to get it to
other trees. This is something where the push had better come from the
other direction.
Because I think it’s stupid. So use somebody else than me to push your
political agendas, please.
Linus
http://thread.gmane.org/gmane.linux.kernel/475654/focus=475721
http://thread.gmane.org/gmane.linux.kernel/475654/focus=475824
“In other words, you guys know my stance. I’ll not fight the combined opinion of other kernel developers, but I sure as hell won’t be the first to merge this, and I sure as hell won’t have _my_ tree be the one that causes this to happen.”
It seems he really didn’t like this idea.
http://thread.gmane.org/gmane.linux.kernel/475654/focus=475721
“In other words, you guys know my stance. I’ll not fight the combined opinion of other kernel developers, but I sure as hell won’t be the first to merge this, and I sure as hell won’t have _my_ tree be the one that causes this to happen”
edit: didnt know someone else would pull the same quote
Edited 2006-12-14 04:55
edit: didnt know someone else would pull the same quote
Ah, but you are Mr. Echo!
What’s funny is, I thought software freedom was one of the greatest values of linux – so why is preventing someone’s software from running, an acceptable way of promoting that? Just because you have ideological differences between those that want FOSS and those that don’t doesn’t give you the excuse to deny others the ability to use whatever software they want.
There’s a big difference between “copy” and “use”. It’s exatcly the same issue whether it’s music or code. You can’t re-distribute other peoples music (becuase it’s _their_ copyright), but they shouldn’t put limits on how you personally _use_ it (because it’s _your_ life).
Same goes for code. Copyright is about _distribution_, not about use. We shouldn’t limit how people use the code.
This is why I love Linus. He understands that ones rights only extend so far as to not impose on other people’s rights.
Oh, well. I realize nobody is likely going to listen to me, and everybody has their opinion set in stone.
And he’s relatively humble about it~
Im glad I dropped Linux for BSD last month.
Im glad I dropped Linux for BSD last month.
We don’t care.
who’s we? the 5 people still using linux 10 years from now?
No, the readers of this thread.
If someone had already dropped Linux last month (and in the case of Manuma I don’t believe it’s actually true), then it wasn’t because of these news. Since the proposed change won’t have an effect for a year, and in the meantime some arrangement will more than likely be reached to accomodate everyone, saying that it was a good thing he switch last month brings absolutely nothing to this discussion. Therefore, any person who understands what’s going on here shouldn’t care about Manuma’s comment.
So I saw a completely useless comment and commented on it. Shoot me.
BTW, there’s no reason to think there’ll be less
people using Linux in 10 years than there is now.
I have great respect for Linus. Unfortunately, he doesn’t have a linux distribution. Redhat, Suse, and Ubuntu do and if they are able to pull off this nonsense, that will be the end for commercial linux development.
Why would they *want* to “pull off this nonsense”…I don’t think they would, at all.
In any case, as others have mentioned, the developers don’t really have a legal leg to stand on here. I have absolutely no doubt that it’ll still be possible to run binary proprietary drivers on Linux in 2008.
That said, it would be nice if NVIDIA and ATI open-sourced their drivers…rumors are that ATI will, perhaps NVIDIA will too.
Bravo for Linus. I respect his commitment to his principles, although personally I wish he’d used the BSD license.
I would remind the folks who developed this idea that many, many years ago, Microsoft put in specific code into Windows 3.1x that would detect the presence of DR-DOS as the underlying OS, and throw out a bogus error message.
There was no technical reason for it.
Microsoft did it to stifle competition. The kernel devs appear to be attempting a religious coup.
I’m pretty sure this violates my own “freedom 0”: The freedom to run the program, for any purpose (freedom 0).
But, by all means, carry on– Doubtless, you have a lot of support from Microsoft for your endeavors.
Linus is the man.
This is stupid. The devs pushing this thing are taking away individual freedoms.
This isn’t about, say, taking a proprietary driver OUT of the official Linux source tree. It’s about removing the end users ability to use whatever drivers they want.
Yes, the NVIDIA drivers are evil because they work so well and they’re closed source. However, the kernel developers want to remove the end users free to do whatever they want with their private copy of the Linux kernel.
how so? you can hack your private copy of the kernel to use non-gpl modules.
Most Linux users are not capable of that, and shouldn’t ahve to be
all this whining about zealots is the same old crowd whining about the same old folks who got us here in the first place. STFU. The zealots got us here and are taking us forward.
FWIW I don’t think everything in the kernel should be GPL. it should be _GPL_compatible_ like the license requires.
“The zealots got us here …” —and are taking us right back to where we started. Thanks for nothing.
About Linus being the man and his “pragmatism”…anyone remember bitkeeper?
About Linus being the man and his “pragmatism”…anyone remember bitkeeper?
Nope, thinking is still the same.
Its bitkeeper stance was move against someone moving the ground under work of others. What Bitkeeper did by their changes would move a lot of ground under people using it.
And here is the same. He doesn’t want to see ground moved under anybody else. He’s taking the same stance for him self as for everybody else. No matter if they are closed drivers writers. I call that fair to the max as he’s not differentiating between him self and others who might not 100% agree with him.
My opinions mostly mirror that of Linus. It mostly seems like I am in the minority, but I am both happy and proud to share it with Linus.
Draconian restrictions in DRM and the likes are bad, but draconian restrictions on anything but GPL are acceptable? I say it is great to believe in open source, but to intentionally lock out and restrict closed source code is hypocrisy.
We demand closed source companies open up their formats to co-exist, and then _intentionally_ lock them out?
What if all the companies pushing closed source said in their EULA’s that you cannot use GPL code to interface with their software? It would be outcry, but that is exactly what the we are doing to them.
Hopefully, this will drive people away from Linux land.
By placing licensing and religious beliefs before the users and their needs they are taking computing backwards.
Sorry but this sort of thing just wont happen. These debates about locking out proprietary drivers have been going on for years, yet nothing changes. But even if it could, FreeBSD/PC-BSD would immediately rise to the occasion.
The reason is simple. It’s completely legal for the *end user* to directly link proprietary modules or even use GPL shims that interface them.
The only thing the GPL has any jursidiction over is *distribution*.
yes and linux absolutely depends on distribution if it wants to be more than a hobby OS for geeks who don’t mind hacking the kernel
I dont’ think you understood Xaero’s point. The kernel devs don’t have a leg to stand on because distribution of a modified kernel that allows non-free drivers and separate distribution of said drivers is perfectly legal. Therefore, even if the kernel devs decide to do this, it will be unenforceable. Things will just continue, exactly as they are right now.
Seriously, I fail to see the problem.
yes and linux absolutely depends on distribution if it wants to be more than a hobby OS for geeks who don’t mind hacking the kernel
I admit I’m a geek but not because I run Linux. I’m more of a geek because I’m single, like computers and high-tech gadgets and gizmos.
Yes, being single sucks sometimes.
The geeks you’re thinking dont use friendly desktop distributions with GUI tools that are reminiscent of Windows.
They use primitive distributions like Slackware and Gentoo and compile optimized kernels for their exact architecture, in hopes of squeezing another 5-10% of performance.
Unfortunately, one thing this decision won’t drive away is uninformed, flamebaiting anti-Linux trolls.
BTW, Bender runs Linux.
… and that’s all I care about, really.
Whenever he chimes in on these issues, I always agree with everything he says. I he’s always on the mark while the other side has an agenda and will “piss on your leg and tell you it’s raining.”
It’s thanks to devs like Linus that we have a functional alternative. He actually cares about the user, rather than just playing lip service. If it weren’t for people like Linus, these folks wouldn’t have anything useful to leverage.
Edited 2006-12-14 06:52
So the end result would turn out
1) reduced users for Linux or
2) reduced users for nVidia and the like.
Clearly the zealots choose software over hardware, but is this case for common desktop user ?
FUD
”
That said, it would be nice if NVIDIA and ATI open-sourced their drivers…rumors are that ATI will, perhaps NVIDIA will too.
”
I’ve heard this on Slashdot too, but what is the source of these rumors?
I’ve heard this on Slashdot too, but what is the source of these rumors?
They’re rumors…who knows where they’re coming from! From what I recall, it’s due to the new management at ATI, who was bought by Linux-friendly AMD.
He seems right to keep the present way, _because_ benevolent meritocracy just makes the world better for everyone as opposed to blind (copyright) law-enforcement.
Somehow he makes the point that copyright-based enforcement of GPL just doesn’t have enough fine moral distinction between what is right, and what is wrong.
BTW That is also a good reason to drop GPLv3.
BTW That is also a good reason to drop GPLv3.
Are you referring to the DRM-thing? v3 doesn’t forbid the use of DRM-measures. It’s just not allowed to be used to restrict the distribution of the sourcecode.
Or which part exactly do you mean? AFAIR the “derived work” clause didn’t change, did it?
But I also think that he is right.
It would change nothing, there would probably just be more userspace programs. Good that we have Linus
If FSF will continue to abuse the users there, I’ll leave linux for any really free platform such as BSD.
Linus truely stands out as the voice of reason among impractical fanatics.
I agree with totally.
I am a hardware guy, if some cretin from the linux crowd tells me I can’t use my shiny and expensive nVidia card because it doesn’t have an open source driver then I will leave linux, after I have verbally kicked him in the nuts for wasting my time over the years.
I *like* linux because I am not arbitrarily restricted in how I use it, if that changes why the hell should I care about it.
Fortunately, I have a great deal of faith in the common sense and pragmatism of Linus.
As much as I disagree with complete enforcement of open-sourcing the drivers, what you just said is plain stupid.
You are saying that because of that, you’d insult people who have WORKED FOR FREE to BRING YOU HIGH QUALITY FREE OS. Now that’s pretty f–ked up man.
NEVER EVER say anything like that. I don’t do OSS and FSF to be f–king told such things. I work for free and anyone who doesn’t like my ways should just go quietly away. You don’t have the f–king right to tell anyone anything unless you payed for it.
If you don’t charge for something, how can anybody pay for it? Especially software? Oh yeah, by supporting you by using the software. If a user doesn’t like what you’re doing, then they will leave, and then, not only are you working for free, but you’re wasting your time if you have no users supporting your ideals.
Everybody is entitled to their opinion, and the whole concept of OSS is the free exchange of source, programs and ideas. Your attitude is wrong, you might not get paid, but when you write OSS software, you either respect your users, or you are irrelevent, as those users are the reason you exist. When XFree86 changed thier liscense, the code was forked and everybody moved to Xorg, and that will happen here too, people will find something that fits thier needs, and then you’ll REALLY be working for nothing
trying to push Linus around.
They should just give it a rest. And give up their silly crusade. The kernel belongs to Linus and devs. He makes the big decisions.
This kind of thing seems to break out and pop up on OS News or Slashdot every couple of weeks at the moment.
The core of the issue is nearly always the same: in order that I can keep my principles pure, I don’t care how much inconvenience I am thereby inflicting on you. I am completely justified because my intentions are beyond question, being entirely good and pure.
The usual word for this is bigotry. No one seems to ask whether it is really sensible to regard every commercial and institutional entity on the planet as intrinsically evil and untrustworthy, or whether these constant attempts to reinvent the wheel are a sane and productive use of resources. Nvidia, for example, has been an outstanding supporter of Linux and produces top-quality drivers. Just what’s so bad about that? No one is forced to use Nvidia’s drivers but that doesnt mean it is right to force people not to use them.
Linux has to exist in the world as it is, as we all do. Those who try to build new, unsullied and Edenic worlds usually end up building prison camps. In this case, anyway, a lot of the potential inmates will see it coming and disappear to Windows, Mac or BSD.
Edited 2006-12-14 08:26
I moved to FreeBSD about a month ago. There is no trace of Linux on my computers any more. Trends in the Linux community reminds me of Taliban ways before they were defeated in 2001. I am running my own business I don’t want my customers to mistaken me for some kind of radical, or political activist. In fact that some of my clients did make some remarks on that in the past.
I decided that it is better to do that sooner rather than later.
DG
I have never enjoyed a topic as much in a long time.
Seriously tone it down with the zealotry comments. Linux would not exist without these people that are so candidly insulted. The majority of software on Linux/BSD/ReacOS etc etc or any of these niche operating systems is GPL, and very little is under any license that does not hold the same values in part or in whole. Even BSD is at the heart about *sharing* code.
I see the arguments like “just work”, “I’m moving to BSD”, and my first thought is that these people do not use Linux. I use linux and from experience I know that every piece of hardware in my machine works with linux, not the other way around. If binary blobs were removed tomorrow. It simply wouldn’t effect me. I know I can buy a *whole* 100% open source linux machine cheaper than I could buy Vista.
Linux’s real market has always been servers. Would they be affected? Not a chance.
@WereCatf How is an dropped driver better than one that is still supported?
>The majority of software on Linux/BSD/ReacOS etc etc or any of these niche operating systems is GPL
It depends. If I don’t use many applications in userspace, there isn’t a lot of GPL arround. Take libc, ls, tar – for example, most of these aren’t from GNU in BSD. Take OpenSSH and many other essential software original developed for BSD. It’s possible to run a BSD system with a very small impact of GPL code (for example Gnu compiler). Many software nowadays GPL’ed is originally derived work from BSD – but you know, GPL is a “viral” license – vice versa it isn’t possible.
>Even BSD is at the heart about *sharing* code.
Yes, because it’s open source – but this isn’t something to emphasize. It’s so since decades, free to *everyone*!
>Linux’s real market has always been servers.
Indeed? This is more the attribute of *BSD, Linux market is more about the vast majority of all kind of multimedia plus server.
>The BSD operating systems are not clones, but open source derivatives of AT&T’s Research UNIX operating system, which is also the ancestor of the modern UNIX System V. This may surprise you. How could that happen when AT&T has never released its code as open source?
It’s true that AT&T UNIX is not open source, and in a copyright sense BSD is very definitely not UNIX, but on the other hand, AT&T has imported sources from other projects, noticeably the Computer Sciences Research Group of the University of California in Berkeley, CA. Starting in 1976, the CSRG started releasing tapes of their software, calling them Berkeley Software Distribution or BSD.
http://www.lemis.com/bsdpaper.html
Remember, people using *BSD tend to be very experienced in Linux. Only a very small amount of people are migrating from Windows to *BSD. And some others never used Linux, only BSD, because some time ago Linux didn’t exist at all
So in the end there are different “worlds”, some that mimics Unix, some that derives from Unix. Linux is “just” the kernel and it’s a very good one, only some of the distros are a mess.
Actually, I use linux quite a bit, I run Ubuntu at home, before that Fedora Core, and I also boot BSD and Windows, but after this, it’s just BSD and windows, as I know where my rights stand with them. I only use Windows at home because I use it at work, and have a test domain setup, otherwise, I wouldn’t use it either.
as far as buying a “linux machine” you are right, but if Linux is ever going to reach out and get the normal users desktop, it has to support them too, and they sure as hell don’t know how to go buy such a “linux machine”
This move on the part of the kernel Devs is no different then the same sort of vendor lockin that MS and Apple do, except here, it locks you into a liscense, one that is just as unbending as any other
“Linux would not exist without these people that are so candidly insulted. ”
Zealot. zealot zealot zealot. I’ll say it again. Zealot.
The zealots give themselves *way* too much credit. What they always fail to realize or admit to themselves is that they didn’t do it single-handed, and that a large portion of the code came from large corporation like IBM and Novell.
You may have started a good philosophy, but stop pretending that you wrote all the code.
Morton was always very reasonable when compared to Linus.
Linus has kept us away from troubles from the fsf though.
its important to note that the reason for this new protocol is to strongarm manufacturers into releasing GPL code,
if Andrew Morton believes that his approach will successfully coerce hardware manufacturers into submission then I’m all for it, ask yourselves has Andrew ever steered us wrong. Lets be civil, calling people fanatics and extremists is savage, we are FOSS inhabitants, we are not from the muck of anti-OSS.
Guys please, i think if we are going to encourage FOSS and FOSS hardware then we have to use questionable tactics.
A FOSS “fanatic” suggested to me sometime ago, that there is a surefire way to move people of windows and all the proprietary nonsense. “let a group be formed thats only function is to create viruses of the standards of sasser and mydoom, then do this once a week and eh voila, people run away from the flaming debris”
i think we need to be more violent and aggressive if we are going to propagate FOSS, I’m buying Intel Motherboards now just because of their open drivers (i used to be an AMD user).
so guys, we need to suddenly pull the plug on manufacturers, then they will feel the heat, when they start to lose cash, thats when they’ll be aching to become FOSS. History has shown us that, when a FOSS project that is popular is launched it is cataclysmic.
look at firefox, and now Intel, watch as their onboard graphics cards have super FOSS drivers built for them, they will be perfectly tuned, now is the time to hit the hardware manufacturers. come on guys has Andrew ever failed us
and please stop this fsf bashing, if you dont agree with their views fine, but you have no right to rudely attack them. i am a KDE zealot and i even i know the FSF is a mostly good entity.
we need another sasser worm, mydoom, every week, to finish windows once and for all
Edited 2006-12-14 09:25
@jango why would you say something so reasonable and well put,and then say something really stupid.
Linus made himself clear about what he believed throughout the Bitkeeper fiasco. He got burnt badly by that one. He must have moved mountains to get Git working.
Linus should keep out of “Politics” or at least making political statements. Its not his strong point. Linus wants for his own reasons binary blobs in the kernel for wider adoption…and he is right, but its a short term view. Making his spat with the FSF personal shows how weak he is.
Edited 2006-12-14 09:39
the reason for this new protocol is to strongarm manufacturers into releasing GPL code
What are you talking about ? Are you serious ? Do you really believe that some 1% share of desktop personal computers can impress manufacturers ? And that is how big Linux share is.
Server market share doesn’t count, because people don’t care about graphics drivers for servers. Plain VGA or VESA is good enough for the server, and no one bothers about graphics acceleration.
Linux drivers for ATI and NVidia are pure courtesy, that will probably go away now.
Face it, you Linux crowd don’t have enough strength to enforce anything.
DG
I’m not really sure what your point is. I agree that that this will not enforce anything AMD and NVidia will work around it, or work with a forked kernel thats more loose about GPL like Ubuntu.
but I think your confused about the 1%. Even 1% when you talk about millions of users is “lots of users”, and I’m sure you aware…its a vocal community(sic).
work with a forked kernel thats more loose about GPL like Ubuntu
I don’t understand that remark. One can not fork GPLed software to make it less GPLed.
Yes, Linux community is very vocal. But 1% is not good enough to lose trade secrets. I don’t think it is about intelectual property, since they are not selling software, but maybe about not revealing the design details to other vendors.
DG
I don’t think it is about intelectual property, since they are not selling software, but maybe about not revealing the design details to other vendors.
I doubt this is really the issue. I have yet to hear anyone from Nvidia or ATI say this.
I guess in theory this could be the case, but then again if both Nvidia and ATI open-source their drivers together, then neither would gain an advantage on the other…
In any case, it didn’t stop Intel from doing it…
then neither would gain an advantage on the other
Maybe they hope to gain advantage on each other. Not all people are defensive. There is a competition out there, and everyone wants a bigger peace of cake, even if that means that the other guy gets smaller one. Yes, we live in Universe where resources are scarce and living beings are fighting over them.
Intel cards are loer performance, they don’t have much to loose.
DG
“Maybe they hope to gain advantage on each other. Not all people are defensive. There is a competition out there, and everyone wants a bigger peace of cake, even if that means that the other guy gets smaller one. Yes, we live in Universe where resources are scarce and living beings are fighting over them.
Intel cards are loer performance, they don’t have much to loose.”
As everyone involved is being investigated for monopolistic behavior its probably not the best thing to discuss “competition”
Intel On-board graphics cards are indeed slower…but they outnumber that of NVidia/AMD’s offering. They have the majority market share. They didn’t just open source their drivers they are actively working on them. For those who fancy stability over speed Intel drivers are unbelievable…and only a year ago. I would have said they were rubbish and advised others to go for a R200 solution. Your right though the did have nothing to loose, they had a lot to gain, good publicity, support from a vocal community, and interest in there chipset for Server offerings. Are they stepping into the graphics market more seriously, absolutely. They are pretty much been forced to by AMD.
If you really need more speed the latest open source R300 run everything Linux gaming can through at it, and is getting better all the time.
To be fair though serious gaming is neither Linux’s strong point, and Linux gamers are in the *Minority*. Gamers in general are in the Minority.
Your right though the [Intel] did have nothing to loose, they had a lot to gain, good publicity, support from a vocal community, and interest in there chipset for Server offerings. Are they stepping into the graphics market more seriously, absolutely. They are pretty much been forced to by AMD.
Certainly, by having open spec’ed cards one of their new 3D cards will be my next one.
There is a competition out there, and everyone wants a bigger peace of cake, even if that means that the other guy gets smaller one.
Actually, sometimes there is more to gain by collaborating instead of competing.
Yes, we live in Universe where resources are scarce and living beings are fighting over them.
Er…I think you’re taking this a bit too far! After all, the first humans flourished by cooperating, not fighting. Then greed kicked in…
Do you really believe that some 1% share of desktop personal computers can impress manufacturers ? And that is how big Linux share is.
Please let’s not start another discussion about Linux market share, m’kay?
In Jan. 2005 it was estimated at around 3%…I doubt it has gone down since then. In any case, it is irrelevant to the discussion at hand. As I and others have noted, the kernel devs will *not* be able to enforce this. Workarounds will be found if necessary.
Also, I think there’s a good chance Nvidia and ATI will open-source their drivers. It would make financial sense, as they would no longer have to maintain them themselves.
In fact, to hardware vendors, drivers are PITA, but again, there are possible trade secrets here.
DG
In fact, to hardware vendors, drivers are PITA, but again, there are possible trade secrets here.
True…however I wonder about the intrinsic value of those secrets vs. the cost of maintaining drivers over an indeterminate length of time…
n fact, to hardware vendors, drivers are PITA, but again, there are possible trade secrets here.
As if they can’t be investigated when they are not opensourced.
so guys, we need to suddenly pull the plug on manufacturers, then they will feel the heat, when they start to lose cash, thats when they’ll be aching to become FOSS.
Actually, I highly doubt that. Have you considered that there’s lots of people using Linux who buy/acquire second-hand hardware? Or just use it on their aging computer? And there aren’t even that many Linux users that hardware manufacturers would feel much..Atleast I can’t see how this’ll force hardware manufacturers to open their drivers. If they can’t do that due to some patent issue, they most likely will drop support for Linux altogether.
@WereCatf you can’t talk believe that you again talking about support for older computers, and closed source drivers…when your r200 had its support dropped.
Read Martin J. Bligh responce to Linus, it covers leverage very well. I suspect Oliver should read it as well. Its good I like this bit the best
“But please can we have the pragmatic argument about what we want to achieve, and why … rather than the legal / religious arguments about licenses? The law is a tool, not an end in itself.”
http://thread.gmane.org/gmane.linux.kernel/475654/focus=475824
Edited 2006-12-14 10:32
Choosing an open source solution. Is rarely a political decision closed source drivers are insecure; badly supported; dropped without warning.
A good example would be the r200 based ati cards under linux, which have had support “dropped” by ati, and yet the open-source drivers will run AIGLX/Beryl which the binary drivers will not run even on there newest cards.
or Nvidia security problems.
http://www.theregister.co.uk/2006/10/17/nvidia_linux_graphics_bug/
Linus is right, it’s about technology not religion.
Quote of the day brought to you by James Morris
“IMHO, it’s up to the users to decide if they want to keep buying hardware which leads to inferior support, less reliability, decreased security and all of the other ills associated with binary drivers. Let them also choose distributions which enact the binary driver policies they agree with.
Linux is precisely not about forcing people to do things.
– James”
You can really tell nobody looked through this damn thread before, making the same remarks.
Who cares if the remarks are repeated, it’s people talking about an important and contraversial topic, we can say what we want, as many times as we want. This is a forum, not a soup line
I don’t think anyone has read Linus’ comment.
He’s made it abundantly clear that he feels that this is a bad thing. Yes, you need to abide by the rules and watch what you distribute with the kernel, but gtting people to confirm with some developers’ world view on what should and shouldn’t be allowed in grey areas.
This is definitely not going into Linus’ tree, and it must first be merged into kernel trees outside in the distributions and tested first before Linus will even consider it.
This is not going into the Linux kernel.
everyone loves freedom. right…
but in a free work , do you have the freedom to enslave others ?! are you free to hurt other people ?!
i would like this to happen , but i dont know if its a good idea.
it is easelly avoidable , a simple patch and its disabled. but at least , its not that grey area anymore.
i really dont know why some ppl say about switching from linux to bsd … i didnt know for example Theo loved binary drivers ( lol )
and i really dont care about users abandoning linux for windows : i just say ” good ridance “.
it linux is to become another “windows”, then please please make this not loading non binary drivers really happen.
i dont mean to sound like an old man , but i remember the days where linux was completly stable.
now ?!? even a sound card or wireless card , or video card , or well , pretty much what ever crashes linux ( linux is actually much more unstable then windows ( but of course now i will read about “i am using nvidia , and i have an uptime of 20 years” )
as long as the patch and kernel remains free and open , this is never ever ever like what microsoft did for dr-dos. i really dont know why we hurt linux just to keep the trolls around.
ps : yes , i know that this is a pretty harsh comment. this is just from someone who remember the days where linux wasnt so popular, and was maybe even better. ( me misses the days of enlightenment , of themes.org , of gnome 1.x ) nowadays its everything for the corporate user. the poweruser ( the one that actually cares about this ) ?! let him screw himself
i really dont know why some ppl say about switching from linux to bsd … i didnt know for example Theo loved binary drivers ( lol )
Theo has certainly campaigned against binary drivers (and actually worked with companies to get free and open specifications), he has left OpenBSD under the permissible BSD license. NVidia is free to ship a version of OpenBSD packed full of binary blobs if it so desires.
The no-binary Linux developers seem to be more interested making a political statement than doing anything practical. If they had half of Theo’s chutzpah, they’d work with NVidia, ATi, et al to get free and open specifications and write drivers themselves. But that requires work. A kernel license patch is easy… and craven.
But that requires work. A kernel license patch is easy… and craven.
Linux has a different position and doesn’t have to beg for driver specs.There’re are numerous big companies who have a interest in linux and in fact use it on day to day basis.
GKH himself has replied in this thread already and admitted, that he (understandably) has been carried away emotionally and had not “the big picture” in mind, when he suggested the blocking of kernel space binary only drivers:
http://article.gmane.org/gmane.linux.kernel/475890
[Note: Somehow, I was not able to edit this comment, when I realized, that I had not given the proper url in the first place. My comment in reply to this one now points to the same url as above.
Weird ]
The proverbial Tempest in a teapot comes to mind.
Edited 2006-12-14 12:07
Oops, posted the wrong url, my appoligies.
http://article.gmane.org/gmane.linux.kernel/475890
is what I meant.
EDIT: Reading again GKH’s post, I don’t get, what he means, with the latter part of this statement:
GPL covers distribution, not usage, no matter how much the people working on v3 want to change that
Is this my sarcasm meter failing, or do the Kernel devs have complains about the GPLv3 shifting from distribtion towards useage? Because, frankly, I’ve not noticed such a thing so far.
Edited 2006-12-14 12:18
I think he is quite clear. He is saying he is too small to hold back, binary anything either in Linux the kernel or Linux + stuff that sits on Linux. However much he dislikes the whole thing technically.
He’s saying without these binary blobs Linux would simply *not function*, and these are becoming more and more common.
as for the v2 vs v3 argument I think some kernel develops wrote a letter some time back its an easy find. I do find it slightly amusing how Linus tries to twist FSF definition of freedom back on itself. I don’t think it works very well.
@trenchsol Feisty is going to ship binary drivers by default. Its big news if you don’t get the Ubuntu reference. Its actually referred to in several of the posts including the one “h times nue equals e” posted
Edited 2006-12-14 12:44
Thanks for your reply. For the record, I’m strongly pro free software drivers and all my current hardware works flawlessy without binary drivers under Linux. Furthermore, (this is in reply to the 1%-market-share guys) I advice my friends/family to buy hardware with open specs and drivers, whenever they ask me for input on their planned hardware purchases, even if they don’t happen to run Linux atm.
Although I have no numbers at hand, I guess that the tech savvy Linux users have quite some impact on the purchase behaviour of their less tech-savvy peers, given, that they tend to be called, when things break.
With that said, I have to confess, that a scheme similar to the one proposed by GKH would make me very uneasy, since it tries to use the GPL (that covers only distribution process) to get a grip on mere useage. This is a situation, where the cause (get rid of binary blobs) doesn’t justify the misuse of the tools (GPLv2), at least imho.
I have interpreted the post by GKH in such a way, that he
– has not thought about all legal consequences of such a prevention mechanism, when he stated, that he would like to ban binary blobs completly within a year and he.
– is very frustrated by the problems binary blobs can cause wrt to system stability and support (here, I deeply sympathize with him, since I have been biten by such things myself already enough)
If I have missed something else between the lines, please feel free to correct me.
Re the discussion v2 vs. v3 : Yes, I know the kernel dev’s position paper [1], and I’ve read it in september when it came out and I’ve read it now, just in case I’ve glitched over something. I’m still not able to identify a passage, where the kernel devs raise concern about a shift from distribution to useage wrt the scope of the GPLv3. They have objections to the DRM clause, the patent clause, the set of additional restrictions, they criticise the drafting process and the role of the FSF, yes. But I wasn’t able to catch GKH’s drift wrt to the useage vs. distribution phrase.
Sorry, if my first post has been unclear.
[1]http://www.linux-watch.com/news/NS5561540450.html
EDIT: Just in case you wondered: I’m also very positive about the GPLv3 and welcome the changes proposed. But I guess that community projects without a “or later” cause will have difficulties to change the license without a required transfer of copyright for all contributions to one entity, no matter how funky the GPLv3 would be. So, although the discussion process itself was very interesting, it was very unlikely in the beginning, that the Linux Kernel would or even could swith to GPLv3 as a whole.
Edited 2006-12-14 13:24
“I do find it slightly amusing how Linus tries to twist FSF definition of freedom back on itself. I don’t think it works very well.”
Huh? It works great, because it’s true. Restricting my usage for the sake of trying to prove that your philosophy is better than mine has nothing to do with software. It’s the mirror image of RIAA.
@Oliver big post, don’t try and and cut and past from the more informed, and misrepresent it. The post you cut from points out what the FSF has contributed to the Linux *kernel* Not Linux/BSD etc or even computing as a whole. I understand the BSD vs GPL licence issue…and understand the benefits/pitfalls of both, and I understand why, but I fail to see the relevance here. the binary blobs are not released under the BSD licence or any other open source licence.
@trenchol true, but GPL 2.0 has had its day hence GPL 3.0. why can you not understand that if there is a patch in Linus’ kernel to remove binary blobs…there cannot be one to put it back in, and it seems pretty trivial. Heard the arguments about trade secrets/Intellectual Property(sic) a long time ago. Its actually mentioned in the thread why its a myth *read it*
It is an issue…and will always be an issue. This is not the first time this has been raised, or the last. Linus will *never* do this. He would need an awful lot of bottle to do this, and Linus wants to be seen as lovably by everyone apart from the FSF. Its a good topic, and particularly in the light of Linus making overtures against FSF and his *users*.
The reality is unless the opensource mid-range graphics cards open-source software are good enough. Its not going to happen.
Its also an Issue because of the 3D-Desktop, currently binary drivers are not keeping up, and the open source ones are weak by comparison.
Edited 2006-12-14 12:14
I am sorry but I don’t understand your point. You seem to reply to my post, but the reply does not seem to be related to it…. Maybe I didn’t make myself clear enough.
DG
Thanks God Linus Torvalds still has something to say when it gets to merging new code in the Linux kernel.
I don’t understand why those who think those modules are illegal don’t go and sue nVidia or ATI. If they are illegal, they’re illegal to distribute, not to load.
But it’s easier to “use” the users to attack those companies instead of going directly against them, right?
I could hope for sooner rather than later but later will do just fine….
The kernel is suppose to be GPL. Glad to see a stand being taken to TRULY make it so!
Edited 2006-12-14 14:47
“I could hope for sooner rather than later but later will do just fine…. The kernel is suppose to be GPL. ”
The kernel is also supposed to work with hardware.
When I read Andrew Morton’s and Greg Kroah-Hartman’s post, I thought: OK, this is the last drop in the glass, I’m switching (to Solaris, to ReactOS, back to FreeBSD or even Windows; it doesn’t matter).
Then I read Linus’ reply. At least someone who has sound reason. And incidentally he’s the leader of this Linux thing. It made me wonder for how long though.
It seems like the majority of Linux developers want this patch to go in and want to enforce the GPL-only modules rule. Is this a good enough reason to fork Linux? And which Linux will distros use when and if the fork happens?
What do you think?
I was hoping that they will fork into some kind of community variant and a consumer variant I was interested in consumer variant, but came to a conclusion that I don’t have time to wait for that to happen.
So I switched to another OS.
DG
but came to a conclusion that I don’t have time to wait for that to happen.
Didn’t have time for what? Even if this makes it into the main tree (and it looks like it won’t), even if the distro makers included the hange (which they probably wouldn’t), this wouldn’t go in effet before 2008…
So I switched to another OS.
I fail to see the rationale behind that. You moved to another OS to protest something that hasn’t happened yet, and may never happen? You’re not making much sense here.
According to this : http://lkml.org/lkml/2006/12/14/63 The patch has been removed by the author and this article is now no longer an issue.
Excellent
To people questioning if this is good enough reason to fork the kernel:
Why would you fork the kernel over 10 lines of code? Most distros have patchset they apply to the kernels anyways, it would be a nobrainer to make a patch that removes the 10 lines in question.
And also, it looks to me like the suggestion to prevent binary drivers from loading isn’t really motivated by politics but more by technical reasons. It all relates to the fact that when you load closed source drivers, debugging the kernel becomes alot harder since you don’t have access to the source.
But ofcorse it also turns into a political stance.
Doing this to prevent the distribution of closed source drivers doesn’t make sense since it’s a completely seperate area. (distribution vs loading), on the other hand, this together with the non-stable binary api will be another thing that makes life harder for those distributing closed source drivers.
The “technical” argument about debugging of the kernel made harder by loaded binary modules is itself politically motivated and not true. First of all you still have access to the source-code of the kernel. What you have no access to is the source of the binary module; but nobody is asking kernel programmers to debug code they didn’t write. Second, you do have access to machine code which is (for me) more helpful in the case of low-level kernel/driver code than C source (even the C source would likely include substantial amount of assembly here and there).
Next we will be trying to force Intel to release the trade secrets inside of their Pentium processors.
This is an ill wind that will not blow any body any good. This will cause most hardware manufactures to back away from even considering releasing a Linux driver for their devices.
Even the Linux server market will suffer seriously if makers of storage devices decide that gains of offering Linux support is not worth sharing their trade secrets.
This is counter productive zealotry that will profit no one. What planet are these guys from? I hope Linus will have the back bone to keep on doing the right thing.
I like the BSD comments. Its normally tied with “just work” etc etc. I was particualarly fond of. “I’m putting my money where my mouth is and buying a MAC”. Good times. Please enough with the melodrama.
The reality is *BSD is pretty good, and there is lots of reasons for choosing it over Linux. Admittedly there are political reasons…but I suspect most of the “I’m moving to BSD” don’t fall into that category.
What I do find most strange. Is I would argue strongly that the distribution, matters much more than the kernel. Or the choice between KDE or Gnome. I know “Linux” the kernel is a poster child for well “Linux”. Its such a small part of it.
@trenchsol really do keep up. Its not time yet to chain yourself to linus’ leg. It was always intended for a year in the future..and was rejected anyway. I suspect your post is a bit melodramatic.
Edited 2006-12-14 15:37
I don’t really see the reason why many posters make a fuzz about something that is only within a few lines of some blog somewhere.I haven’t seen a link to any official statement regarding the issue.Linus has been right many times before.
That’s like comparing Ghandi to Hitler because they were both vegetarian.
Enforcing the GPL and its spirit is a good thing.
GPL is about copylefts, not copyrights.
The FSF is not the RIAA.
Edited 2006-12-14 16:28
I could hope for sooner rather than later but later will do just fine. The kernel is suppose to be GPL. Glad to see a stand being taken to TRULY make it GPL compliant!
As I have said in the past, I think it is time for a split. Let the “open source” guys have their kernel and let the “free software” guys have their kernel. A nice amicable split that stops the in-fighting, provides some competition and provides a choice for the kernel.
Or solaris can go v3 and we can all switch!
“””
Let the “open source” guys have their kernel and let the “free software” guys have their kernel.
“””
The free software guys *do* have their kernel. It’s called the Hurd, and is a great example of what happens when you start basing your technical decisions on politics instead of technical merit. (Yes, I know it’s not strictly correct to call it a kernel so let’s not go there, guys.)
I’m quite happy with that split, and very happy, indeed, that Linus has had the good judgement to steer us clear of that particular danger… so far.
The free software guys *do* have their kernel. It’s called the Hurd, and is a great example of what happens when you start basing your technical decisions on politics instead of technical merit. (Yes, I know it’s not strictly correct to call it a kernel so let’s not go there, guys.)
The HURD has nothing to do with this. Technical decisions are what has slowed the development of the HURD along with not enough actual people working on it.
The linux kernel is covered by the GPL therefore it IS free software. The problem is that violations can only be addressed by copyright holders so as long as the copyright holders do not care about a violation then nothing can be done.
Edited 2006-12-14 17:32
“””
The linux kernel is covered by the GPL therefore it IS free software. The problem is that violations can only be addressed by copyright holders so as long as the copyright holders do not care about a violation then nothing can be done.
“””
Yes. Because it is *their* copyright on *their* hard work. And *they* have the privilege of deciding what they want to enforce and what they prefer not to enforce.
All too often, some in the community seem to feel that the control of the copyright really *should* be theirs, or that the actual copyright holder has some sort of responsibility to them. Often, simply because they use the copyrighted work, and evangelize for Free Software.
Yes. Because it is *their* copyright on *their* hard work. And *they* have the privilege of deciding what they want to enforce and what they prefer not to enforce.
All too often, some in the community seem to feel that the control of the copyright really *should* be theirs, or that the actual copyright holder has some sort of responsibility to them. Often, simply because they use the copyrighted work, and evangelize for Free Software.
Don’t try to turn this into control. Has nothing to do with control.
Why bother licensing under the GPL if you aren’t going to follow it?
If anything it would seem the GPL was chosen for the linux kernel to get people to contribute to the kernel, make it popular and then just do whatever they want to with it.
And this is EXACTLY where a fork should come in… as I talked about one for the guys who do not care about the violations and one for those who do. Nice and clean.
Edited 2006-12-14 17:48
“””
And this is EXACTLY where a fork should come in… as I talked about one for the guys who do not care about the violations and one for those who do. Nice and clean.
“””
You are actually advocating a fork in Linux compatibility at not just the binary api level, but the source code level?
All the nightmare scenarios about Linux fragmenting “Like Unix Did” would be validated. Rightfully validated. As in “not just FUD anymore”, but the actual truth.
I think you are either out of your mind or haven’t thought this through.
And in case anyone is wondering… *this* (forking the kernel) is the threshold at which I start taking NotParker’s “cultist” view of us more seriously.
Fortunately, there will be no fork of any significance.
Edited 2006-12-14 18:07
And in case anyone is wondering… *this* (forking the kernel) is the threshold at which I start taking NotParker’s “cultist” view of us more seriously.
Meme: http://en.wikipedia.org/wiki/Meme
“””
Meme: http://en.wikipedia.org/wiki/Meme
“””
Don’t push it, Bruce.
Actually, I’ve read “The Selfish Gene” and its follow up (the title escapes me at the moment) for myself. Note that I was careful to put your term in quotes.
Like most OSS advocates (whom you probably hear less from than from those with more extreme opinions) I have no desire to join a cult; least of all your “Cult of One”, committed to attacking anything that is not your OS of choice.
Correct me if I’m wrong, but I believe that I have noted your attacking, at various times, Apple advocates, OSS/Linux advocates, RISCOS advocates, and Haiku advocates. Not sure about the BSDs, SkyOS, and ReactOS. However, given your Xenophobic attitudes, I imagine I’d find those attacks if I looked through the archives, which I don’t care to do.
But if there is a meme being created and strengthened here, it would more likely be in reference to your own “Xenophobic Cult of One”.
Hey, I think I’ll start using that more frequently. How does “XCoO” strike you? Even starts with a capital X, which is always popular in tech circles.
Like most OSS advocates (whom you probably hear less from than from those with more extreme opinions) I have no desire to join a cult; least of all your “Cult of One”, committed to attacking anything that is not your OS of choice.
I’m not committed to attacking anything except those who act like a cult and those who hate Microsoft and Windows in a dishonest cult-like manner.
I could list the rest of the dishonesty that annoys me … but it would take too long.
I look back on my first comment which was a snotty response to what seemed like an unending series of attacks on Windows.
http://www.osnews.com/permalink.php?news_id=14774&comment_id=129566
I think the commenters I was responding to deserved it. Of course I got modded down and attacked for it.
And I’ll keep pointing out the cult-like behavior of FOSS advocates who do spend a lot of time attacking each other and former members of the chosen (like Novell and Miguel Icaza) with cult-like vigor but reserve most of their hate for Microsoft.
(I actually prefer OpenVMS, but thanks to the sale of Digital and Compaq it isn’t going anwhere these days)
Well, I guess we’re pretty off-topic. But I think its a worthy subthread even so.
(BTW, as I’m sure you know, good security is like good hiking gear. Multiple layers are best, and a firewall is a good first layer. But I imagine we already agree on that.)
NotParker, your posting style is absolutely guaranteed to get you modded down, no matter what group you are attacking, except maybe Quakers.
I’m not saying its right, just that you shouldn’t be at all surprised. (BTW, I +1’d you on this post because I think it deserves it.)
As I’ve said before, I think you could be more effective by simply changing the *tone* of your posts. I’m *not* saying that you should bite your tongue and not say what you think. Just that a *little* bit of conciliatory tone would help a lot.
I know I’ve said this before. And I hope you don’t feel harped on. Please consider this as a sincere constructive criticism.
A confrontational tone automatically puts people on their guard, even pushing them into a corner where they might argue a position that is more extreme than they would normally. (It happens. It’s happened to me, and in the middle of arguing back, I realize that I’ve been pushed into defending and opinion which is more extreme than I normally would.)
I think we probably have a sort of middle ground where we could agree.
We both feel that some people have a more extreme “Stallmanist” leaning (for lack of a better term) than we care for. You think they are cultist fundamentalists. I tend to feel that they are well meaning but simply a bit overzealous.
We can probably both agree that extremism is a less effective strategy than one that allows for a certain amount of mutual compromise.
I think that OSS is a good solid development methodology that is better for some problems than for others, and has a role that is expanding as connectivity between programmers around the world increases. We’ve done great with infrastructure and computing “staples”. We’ve done less well (*much* less well) in areas like business accounting. In fact, OSS is pretty much nonexistent in that area, as well I know. I hope that someday it will have a stronger presence there.
And yes, I have a “Stallmanist” side, as well. I really *like* the philosophy. But I have to temper that. (And no, I don’t find proprietary software to be immoral. Often disadvantageous. And I’m much more comfortable with OSS software. But, in my opinion, it’s silly to argue that it’s immoral.)
I promote OSS. But I’m too immersed in the “Real World” to be very extremist. My customers are businesses. And I have to offer them business reasons for using OSS. If I go in and give a speech about software freedom that sounds like Oliver Wendell Douglas’ speeches about the American Farmer, I get looked at like Oliver Douglas did. (“Green Acres” reference in case you are not familiar with it.) It hurts my credibility if I do not stick to business and business facts.
About Microsoft. I dislike them. I dislike them quite a lot. And honestly, I’m probably using a less strong term than is the actual truth.
It’s a complex issue, but I think that one reason that I feel as strongly on the matter as I do is that for the 18 years that I have done support professionally, Microsoft has been in the way of my implementing the solutions that I have felt would be best and most effective. (I’ve long been a champion of Unix and Unix-like OSes… years before the first line of the Linux kernel was written.)
It wouldn’t be so bad if MS had simply had the superior solution, but, for example, even way back 18 years ago, I had to convince customers that essentially wanted to use DOS as a server that I had something better.
Even you wouldn’t argue for the technical excellence of
DOS over Unix. (At least I don’t think you would.)
What MS had was marketing and mindshare.
I *knew* what I had was better for the things we were proposing it for. And Unix’s for 386 were *not* expensive. I’m sure that Sun’s and SGI’s solutions were, but that’s not what I was proposing. (More like AT&T Unix 386.)
After 18 years of an uphill battle, not against technical excellence, but against mindshare and marketing, I do have some pretty negative feelings which I think are simply to be expected.
Note that the technical aspects of MS products have improved radically since DOS, and I don’t really want to get any sort of modern comparison other than to say that I see places where one does better than the other and vice versa, and areas where they both suck and something else is a far superior solution.
Thing is, for all the jokes about “World Domination”, I would be unhappy with a world where 90% of machines ran Linux. Not as unhappy as in the present situation where they run Windows, but still unhappy. (And likely I would be using something else.)
I’ve always kind of figured that we would be best off with something like 33% Linux, 33% Windows, and 34% a mix of other stuff, with the BSD’s making a strong showing there. (Those numbers are negotiable.) And in that sort of world, *everyone* would have incentive to be interoperable. Not philosophical incentive, but actual *business* incentive.
In a world like that, I think we would all be better off. Users of what we call *alternative* OSes would obviously be better off. But users of Windows might actually benefit most of all. No company performs at its best, or has the customer support that it could, without *real*, healthy competition.
It’s not going to happen next year. Or in the next five years. But maybe in 10-20 years the world can look a lot more like what I am describing.
And if I can be part of making that happen, in my own small way, today, in 2006… I’m proud of it.
Sorry for the long, rambling, off-topic post.
Edited 2006-12-15 00:27
NotParker, your posting style is absolutely guaranteed to get you modded down, no matter what group you are attacking, except maybe Quakers.
Its been molded by the attack-dog take no prisoners hate filled style of the cult.
But thanks for your post. I happen to think sensible people are in the minority on the FOSS side since hate for Microsoft and Windows and those who use Microsoft software by choice is never modded down on forums like this. Its encouraged most of the time.
Edited 2006-12-15 00:27
BTW, I forgot to mention that while I’ve not used VMS except for a little bit in college many moons ago, it’s good to hear you like it. From what I’ve heard, it’s always sounded very well designed. As you say, though, fate has not been kind.
Perhaps you could write something about it for OSAlert?
Just a thought.
Its been molded by the attack-dog take no prisoners hate filled style of the cult.
It’s nice to see you didn’t understand a single thing that he said.
*sigh*
I’ll repeat it again, because patience is a virtue: I don’t hate Microsoft. I don’t hate Windows. I make my living producing proprietary software. I try to respond thoughtfully and in a civil manner to people who often don’t feel like doing the same.
Stop trying to demonize those you disagree with, NotParker. Stop insulting those who mod you down when you are off-topic and/or abusive – or even when you get modded down for no reason at all. It happens. Let it go.
Try to have an open attitude. Try to recognize that Microsoft has not always played nice with the open-source movement. Try to understand why people are wary of quasi-monopolies.
Confrontation solves nothing. We should be here for the exchange of ideas, not the butting of heads. Sure, some people may post flamebait towards Microsoft, a company that is obviously close to your heart – but by responding in kind, you are only letting those flamebaiters control how you react.
I’m willing to extend an olive branch to you. Let’s agree to disagree, and try to maintain a civil dialog. I’ll stop calling you a troll, and you’ll stop referring to FOSS advocates as if they were all members of the same religious cult. After all, if there’s a group of people that doesn’t all think alike, it is FOSS developers/advocates – the story at the top of this thread is a shining example of that.
Sure, there are misgivings on both sides, but just like Israelis and Palestinians, we are forced to live together, and therefore we must find a common ground if peace is to exist one day.
What do you say?
I don’t hate Microsoft. In fact i think they will one day be a big supporter of open source, like IBM is today. Remember that once upon a time, IBM used to be the big bad monopoly and every other PC was just an IBM PC clone.
Open source should be compelling by its merits. It shouldn’t be forced by politicians and lawyers as if we are insecure about its potential to change the way we look at software. Ultimately it is the most adaptable, most flexible solution and it’s going to win.
It’s not easy taking the long term perspective.
Open source should be compelling by its merits. It shouldn’t be forced by politicians and lawyers as if we are insecure about its potential to change the way we look at software.
Thats a good point. Lots of people and especially Microsoft competitors are attempting to legislate the use of FOSS.
I’d have more respect if merit and functionality was the decider int he competition.
But I think IBM and Google and other are using FOSS as a proxy to attack Microsoft and have no interest in FOSS as a philosophy … just as a weapon.
[/i]Ultimately it is the most adaptable, most flexible solution and it’s going to win.[/i]
I think Microsft is a moving target that has improved its product a great deal over the last few years. I don’t think FOSS will win anytime soon.
Thats a good point. Lots of people and especially Microsoft competitors are attempting to legislate the use of FOSS.
Only governments can legislate. And it’s their perogative to mandate open standards for official documents. It’s also their right to switch to FOSS software if they choose so.
Tell me, when a goverment decides to buy Microsoft, is that “legislating closed-source software”? Why do you claim to be about choice when you deride anyone who dares to choose something other than Microsoft?
archiesteel: “I try to respond thoughtfully and in a civil manner to people who often don’t feel like doing the same.”
Thats completely false.
Edited 2006-12-15 03:33
“””
archiesteel: “I try to respond thoughtfully and in a civil manner to people who often don’t feel like doing the same.”
Thats completely false.
“””
Well, the post you are replying to seems pretty civil and thoughtful. To me, anyway.
At any rate, I’ve been thinking about your statement that you feel that reasonable OSS advocates are in the minority.
It makes sense, I think, to specify the demographic.
I mean, there are OSS users who just use the software without having any particular philosophical attachment. (Many of my customers fall into this category.)
Then there are the “armchair evangelists”. People who’ve not written any actual OSS code to license under any license.
Then we have people who do professional support and use OSS in their work and implement it for their customers or employer.
And then we have people who actually write the code and decide on the licensing. Or write the code and the company paying them to do it decides the license, or whatever.
Obviously, there is going to be a great deal of variation and/or overlap within each of these groups. But it seems to me that the general attitude in each group is likely to differ from the others.
I’d be interested in your views along these lines.
In particular, would you say that OSS people you would consider reasonable would be in the minority among developers?
Linus made an interesting point in his commentary about that letter and poll regarding GPLv3 that a subset of the kernel developers wrote a while back. He commented that most of the noise regarding GPLv3 was being made by people who didn’t actually have any code to be licensed under it. I have a sneaking feeling he’s right.
(Just in the interest of full disclosure, I advocate OSS, use it in my professional support work, and have made no actual OSS code contributions of any significance.)
Edited 2006-12-15 04:38
Well, the post you are replying to seems pretty civil and thoughtful. To me, anyway.
Ask him how many hundred comments he’s modded down just because they don’t express the right level of hatred towards all things Microsoft or the references say something he doesn’t want to hear.
In particular, would you say that OSS people you would consider reasonable would be in the minority among developers?
Linus made an interesting point in his commentary about that letter and poll regarding GPLv3 that a subset of the kernel developers wrote a while back. He commented that most of the noise regarding GPLv3 was being made by people who didn’t actually have any code to be licensed under it. I have a sneaking feeling he’s right.
For years there has been a cult around FOSS that is common that I will call the “holier than thou”. They compete to be more pure. And those that are’t pure enough are detested. There is always a Devil of some kind around — in this case Microsoft.
Over the last few weeks it have been Novell and mono and Miguel Icaza.
Everything I’ve read from Linux Torvalds has been sane and pragmatic. So I find it confusing that so many hateful people have coalesced around FOSS. I assume a lot of it has to do with the early beginnings of Stallmans religion which seem to be based on getting revenge on those who wronged him in some early project.
They also seem to really get off on trying to bully people into believing and saying what they think is right.
They also tend to be remarkably thin-skinned. They love to rant on and on about Microsoft and Windows … but if you say one snotty or sarcastic thing (despite it being factual) about FOSS they go insane with rage.
archisteel epitomizes that fanatacism. He imagines himself the OSSnews police with the duty to mod down everything that annoys him. Coming to his defense lowers my estimation of you.
Ask him how many hundred comments he’s modded down just because they don’t express the right level of hatred towards all things Microsoft or the references say something he doesn’t want to hear.
None, actually. I only mod posts down when they are off-topic and/or abusive.
archisteel epitomizes that fanatacism.
The only fanatic here is you, NotParker. You have insulted me, ignored my arguments, twisted facts and figures to support your agenda, trolled Linux threads, lashed out at anyone who dares to cricitize Microsoft, and now you have refused my plain and honest offer for a more civil debate.
Stop trying to win debates by attacking other people’s credibility, and start discussing the issues *honestly* and *politely*. That includes stopping trying to portray FOSS users, developers and advocates as this monolithic religious-like group. If I was such a Stallman-worshipper, as you try to portray anyone who dares to question your talking points, then how come I make proprietary software for a living? For your information, I’ve actually made money for Microsoft. I have a Xbox, too, and plenty of games for it. I might very well buy an Xbox 360.
For the last time, I don’t *hate* Microsoft, I just don’t *trust* it with regards to FOSS…now, if it starts to finally behave in a civil manner towards us, then the trust in them will gradually return. I’ve told you this many times, but you keep ignoring it because you *need* to paint anyone who disagrees with you with the same large brush. The problem, for you at least, is that you’re not fooling anyone.
This is what happens when someone holds out an olive branch to you? You take it and slap them in the face with it?
I have been civil. I have patiently repeated arguments which you have ignored. I have let you insult me without lashing back. It is true I have called you a troll, but only when you have, in fact, trolled – that is to say made provocative remarks in order to elicit emotional responses. You yourself have admitted to trolling (I remember a post where you said that people modded you down anyway, so you might as well call them cultists), but even without these admissions one can easily go through your list of messages to verify this claim.
In any case, it’s good that you have decided to show your true colors once and for all.
My offer of civil debate, of course, is always available, but somehow I don’t think you’ll take it…
Dear NotParker,
Your extreme fanaticism, propaganda, and hatred towards all things FLOSS is unmatched by even the most prominent Microsoft opposers on the internet today.
Quite simply, you make “the cult” look like a group of angels in comparison to whoever you represent.
Keep this in mind as you continue preaching your propaganda, insults, and lies upon us.
Sincerely,
Vincent H.
Your extreme fanaticism, propaganda, and hatred towards all things FLOSS is unmatched by even the most prominent Microsoft opposers on the internet today.
Get serious. I’ve never even come close to the fanaticism displayed by FOSS supporters in this topic today … let alone the vitriol and hate they’ve dumped on Novell, SUSE and Migeul Icaza in the last couple of weeks.
You are a joke.
A very thoughtful post, and very well articulated. I’m just afraid that it’ll fall on deaf ears…
You are actually advocating a fork in Linux compatibility at not just the binary api level, but the source code level?
All the nightmare scenarios about Linux fragmenting “Like Unix Did” would be validated. Rightfully validated. As in “not just FUD anymore”, but the actual truth.
I think you are either out of your mind or haven’t thought this through.
Unix was/is open source. Linux is free software. No truly ‘ugly’ fork can occur. That is one of the great things about GPL free software is that a “fork” never means a true splinter or fracture. Open source has no such protection.
I am out of my mind I will give you that one. But I would ask you to think it through since we are talking about free software and not just open source software.
If anything it would seem the GPL was chosen for the linux kernel to get people to contribute to the kernel, make it popular and then just do whatever they want to with it.
Don’t confuse the issue, it’s not about the developers doing whatever they want, it’s about the limits on what they can do. The GPL inherently prohibits the developers from deciding what people get to do with the product and how they use it. That’s one of the Four Freedoms(tm) you’re very fond of.
And this is EXACTLY where a fork should come in… as I talked about one for the guys who do not care about the violations and one for those who do. Nice and clean.
Why a fork? The kernel delivered by kernel.org is 100% GPL. Your real issue is with the distro packagers who don’t feel as bound to FSF ideology.
If you want a pure GPL kernel, it’s simple, don’t use binary modules. Stick with GNU’s or similar minded distros. No fork required.
The attack on “binary blobs” isn’t really about ATI/Nvidia.
Its about Microsoft.
Someday (maybe 10 or 20 years from now) Linux may have 3 or 4% of the market. Or it may have zero.
But, if it does hit 3% or 4%, Microsoft may want to port some of its products to businesses that insist on using Linux (no real evdience of it yet).
The most logical way of doing this is via a “binary blob” of Office 2014 or 2021.
If you stand with Linux, you stand for Microsofts right to use binary blobs for its products.
If you don’t, you are against Microsoft and for purity.
Looks like no one is feeding the trolls today.
“””
The attack on “binary blobs” isn’t really about ATI/Nvidia.
Its about…
“””
Please try to stay on topic. Not all threads have to be about your favorite OS and the company behind it. Deal with it.
It’s a grand notion but like all projects sooner or later reality will rear it’s ugly head and make the final decision as to whether it will be for the good or detriment of it. Cooler heads will prevail and from the sound of Mr. Torvalds last post it seems they already have.
What part of the following comment violates the rules on OSAlert worthy of it dropping to -2 in about 60 seconds?
“The attack on “binary blobs” isn’t really about ATI/Nvidia.
Its about Microsoft.
Someday (maybe 10 or 20 years from now) Linux may have 3 or 4% of the market. Or it may have zero.
But, if it does hit 3% or 4%, Microsoft may want to port some of its products to businesses that insist on using Linux (no real evdience of it yet).
The most logical way of doing this is via a “binary blob” of Office 2014 or 2021.
If you stand with Linux, you stand for Microsofts right to use binary blobs for its products.
If you don’t, you are against Microsoft and for purity.”
Dude, you’re being targeted by a downmod stalker — somebody, apparently, with a very low sense of self-esteem.
Actually, I have to agree at least in principle with a small part of your comment
“The attack on “binary blobs” isn’t really about ATI/Nvidia.”
Precisely, although I would keep the word “attack” out of the sentence. If you dare to read the response by GKH later on the lkml thread[1] (the one where he adresses the removal of the patch, just in case nobody else has noticed), you may stumble across this passage:
[side diversion, it’s not the video drivers that
really matter here everyone, those are just so obvious.
It’s the hundreds of other blatantly infringing binary kernel modules out there that really matter. The ones that control filesystems, cluster interconnects, disk arrays, media codecs, and a whole host of custom hardware. That’s the real problem that Linux faces now and will only get worse in the future. It’s not two stupid little video drivers, I could honestly care less
about them…]
So although NVIDIA and ATI are the two most prominent (and most widely discussed) vendors with binary blobs, they are neither the only ones and (at least according to GKH) not the ones with the single most pressing issues.
Besides this (probably unintended, who knows) insight, I have to disagree with your statement. For starting, I guess you have a sound theory why Microsoft would need binary blobs to actually port their office product to competing plattforms, when all other office suites need little to none hooks into kernel space. But since this discussion is about the (now removed) proposed kernel patch for binary kernel modules and not about porting Office 2014 onto Linux Kernel 3.6 in 10+ years from now, I would suggest we keep the discusion on topic, shall we ?
[1]http://article.gmane.org/gmane.linux.kernel/475890
Edited 2006-12-14 17:52
Besides this (probably unintended, who knows) insight, I have to disagree with your statement. For starting, I guess you have a sound theory why Microsoft would need binary blobs to actually port their office product to competing plattforms, when all other office suites need little to none hooks into kernel space.
Microsoft doesn’t want to open source their Office code anymore than ATI and Nvidia want to open source their drivers.
Binary blobs would be the logical way to “port” Office to Linux (if hell actually froze over).
I’m really curious.
For those of you who are on Torvalds side on this issue, if it was Microsoft using binary blobs, would that change your mind?
And, the other way around, if you are against binary blobs, but you had a chance to run Office or VB.Net on Linux via binary blobs, would that change your mind?
Ok, sorry if this is news for you, but:
You are allowed to use binary-only user space applications on GNU/X/KDE/GNOME/whatever/Linux systems, given you respect the license for all other parts of the system. Pray tell, how is/was it possible to use Word Perfect (binary, closed source) on a Linux System without binary blobs (I know it is possible, because I did so in 1998 – 2000)?
How is it possible to use Acrobat Reader (a favourite application of both of us, I know), again a binary, closed source application, on such systems without binary blobs, even today?
kernel space != user space
aggregation != static or dynamic linking
(aggregation is OK with the GPL, even for binary, closed source applications)
So how is it that MS would need a binary kernel driver to port a pure userspace application like MS Office ?
Edited 2006-12-14 18:02
So how is it that MS would need a binary kernel driver to port a pure userspace application like MS Office ?
Is Office a “pure userspace” application?
It would depend on how much of the underlying WIndows kernel is needed to run office.
It may be that some parts of the Window kernel are needed for the office port to work. The kernel space binary blob would be needed to make Office run.
Is Office a “pure userspace” application?
I believe it is. There’s no reason why any of it should run in kernelspace (and that would be very dangerous anyway).
It may be that some parts of the Window kernel are needed for the office port to work. The kernel space binary blob would be needed to make Office run.
I find that very doubtful. Do you have any sources to support your assertion that parts of Office run in kernelspace?
Is Office a “pure userspace” application?
It would depend on how much of the underlying WIndows kernel is needed to run office.
It may be that some parts of the Window kernel are needed for the office port to work. The kernel space binary blob would be needed to make Office run.
The amount of nonsense that you can post non-stop is truly mind-boggling. One cannot help to think that you are paid to spout this BS as part of a large propaganda campaign.
For your information, Microsoft Office 2000, XP and 2003 already run in user-space on Linux. It’s called Wine/Crossover office. Wine is an implementation of the Win32 API on Unix/Linux.
Of course, Microsoft would love to use DRM and TPM chips on motherboards to tie specific versions of Office to specific computers and specific releases of its operating system, but these decisions would have no technical basis on which to stand.
Of course, it could use kernel hooks to do the above, but that would be of its own volition for as much as I despise DRM, it can be implemented in user-space (search for public/private key crypto systems).
Did a penguin bite you at the zoo? Because otherwise your obsession is pretty hard to understand unless as I stated earlier, you are just doing your job and being paid for the crap you post here.
For your information, Microsoft Office 2000, XP and 2003 already run in user-space on Linux. It’s called Wine/Crossover office.
I thought Access 2003 didn’t work and Access 2000 only worked with a ton of bugs.
And the rest of the 2003 apps only partially work.
And Crossover isn’t exactly free.
Of course, Microsoft would love to use DRM and TPM chips on motherboards to tie specific versions of Office to specific computers and specific releases of its operating system, but these decisions would have no technical basis on which to stand.
Of course, it could use kernel hooks to do the above, but that would be of its own volition for as much as I despise DRM, it can be implemented in user-space (search for public/private key crypto systems).
So they could use binary blobs in kernel space.
Thats what I thought.
Did a penguin bite you at the zoo?
No. You?
Because otherwise your obsession is pretty hard to understand unless as I stated earlier, you are just doing your job and being paid for the crap you post here.
You people always fall back on the “paid shill” accusation.
Don’t you ever wonder if people do things for free out of principal? Is it always about money with you guys?
I haven’t used Access 2003, but Access 2000 works now as well as it has ever worked on Windows, at least in my experience.
Nobody I know uses Access 2003 as it becomes a pain to share files, since the majority of the installed base for Microsoft Office appears to have settled on the Office 2000 formats as the default, irrespective of which version they use.
You did not answer my point at all. You do not need any form of binary blob, let alone have it run in kernel space, for an office suite to run on another operating system, but you can go on making crap like this for as long as you would like to do so, Mr. NotParker.
Nobody I know uses Access 2003
Obviously not since it doesn’t run on Linux. And you left out the list of other Office 2003 apps that don’t run properly in Crossover.
By the way Access 2000/XP/2003 all come with a runtime version of SQL Server (an excellent database).
You probably need some of the Windows Kernel to make that run on Linux.
You did not answer my point at all.
I think I did. I’m guessing you didn’t understand it.
You do not need any form of binary blob, let alone have it run in kernel space, for an office suite to run on another operating system,
You seem to need something extra to get Office 2003 working. And SQL Server.
but you can go on making crap like this for as long as you would like to do so, Mr. NotParker.
I don’t think its crap to discuss issues politely.
You seem to need to be impolite.
You can use Office 2003 without an SQL server. Just the standard built-in jet engine will do fine, which is how most people that use Access use it. If they need a serious database, they might use SQL-server with Access for data entry and reporting.
To get back on topic, you do not need an SQL server to run in kernel space and god knows why you would claim othewise.
I assure you that MySQL, Posgresql and a host of other databases do not run in kernel space.
There is no sound technical reason for such a thing. Therefore, should Microsoft choose to offer Office or any other product on top of Linux, there wouldn’t be any sound technical reason for needing a binary blob in the kernel.
Edited 2006-12-15 04:10
“By the way Access 2000/XP/2003 all come with a runtime version of SQL Server (an excellent database).
You probably need some of the Windows Kernel to make that run on Linux.”
Actually no. You need some dll’s, but none of the kernel in actuality. Crossover Office does support Access 2000 quite well in it’s entirety.
You make some good comments here and there, may I suggest maybe wording them a bit differently or trying things actually? I use whatever OS is the right tool for the job myself, be that Windows or something else. I do think you have not tried one of the alternatives with some of your comments, but then you also may have had bad experiences with them as well.
Actually no. You need some dll’s, but none of the kernel in actuality. Crossover Office does support Access 2000 quite well in it’s entirety.
But not the MSDE which is an optional component of Access 2000. And not Access 2003.
Access as front end plus SQL Server (or MSDE) is a great combination. I’ve written a lot of small/medium application using those two products.
I wouldn’t want to go back to using Access as the backend. I like SQL Server.
Perhaps I’m a puritist, but on a monolithic kernel like linux (where all kernel applications are executed in a commen adress space) userland applications should be seperated as much as possible from kernel space (This is already too much off-topic, so don’t let this turn into another microkernel vs. monolithic kernel vs. hybrid kernel debate, thanks). That’s one of the reasons why we have many different virtual file systems (proc, sysfs, …), deamons, hardware- and signalabstraction layers, the whole POSIX layer, etc. to allow a better seperation/abstraction of userspace applications from kernel applications and kernel implementations.
What Office Suite typical functions are so tightly bound to hardware, that it wouldn’t be possible to use an abstraction interface instead? If vendors can port whole applications by linking against lib-wine, which is afaik under the LGPL (again, please correct me if I’m wrong on this), why wouldn’t it be possible to do the same on behalf of MS Office, if Microsoft would ever decide to do so?
Finally, why was it possible to port MS Office over to Mac OS without kernel patches ?
EDIT: Fixed some typos. I also appears, that I’ve replied to the wrong comment, so this should really be [6]… instead of [5]… . My appologies
Edited 2006-12-14 18:42
Finally, why was it possible to port MS Office over to Mac OS without kernel patches ?
Excel was developed first for the Mac in 1985 and did not get ported over to Windows until 1987. (Bet you didn’t know that)
And Word started as “Multi Tool Word” on Microsofts version of Unix: Xenix.
Word was then ported to DOS and then to the Mac in 1985.
But I don’t know how tied current code is to the Windows kernel, but I do know that Outlook and Access do not exist in the Mac version. Access for sure has deep ties to Windows as does Outlook to a lesser degree.
What Office Suite typical functions are so tightly bound to hardware, that it wouldn’t be possible to use an abstraction interface instead?
I don’t know.
But it may be that Microsoft would use “binary blobs” just to ensure that Office ran on only one distro: Microsoft Linux.
Who would care if they did?
Can I see a show of hands of people who’d actually use Microsoft linux?
I, for one, certainly wouldn’t touch it with a 10 foot pole. At least, not until it was the only OS division in the company.
Edited 2006-12-14 18:57
Another -2. And I’m being good. I haven’t even used the c*** word yet. Informative. Still talking about binary blobs.
WOW.
Finally, why was it possible to port MS Office over to Mac OS without kernel patches ?
Excel was developed first for the Mac in 1985 and did not get ported over to Windows until 1987. (Bet you didn’t know that)
And Word started as “Multi Tool Word” on Microsofts version of Unix: Xenix.
Word was then ported to DOS and then to the Mac in 1985.
But I don’t know how tied current code is to the Windows kernel, but I do know that Outlook and Access do not exist in the Mac version. Access for sure has deep ties to Windows as does Outlook to a lesser degree.
What Office Suite typical functions are so tightly bound to hardware, that it wouldn’t be possible to use an abstraction interface instead?
I don’t know.
But it may be that Microsoft would use “binary blobs” just to ensure that Office ran on only one distro: Microsoft Linux.
Well, they can always try to do that :-). I don’t know whether Microsoft, Microsoft Office, Linux or the seperation between kernel space or userspace will be around or even relevant in ten years or so (probably yes, inertia is a funny thing).
But I stand to my point, that the MS Office can’t be tied that close to the MS Windows kernel, since according to[1] Microsoft Office 2003, XP, 2000 and 97 (including outlook for all mentioned versions and access for Office 2000) are already supported by CodeWeavers crossover package. Which, I would like to add, is based on the reverse engeneered Windows API (WINE) and is as it seems rather independent of a special linux kernel version[2].
If MS office developers can’t port an application to which they have full source-code and documentation access with an abstraction layer like winelib when other parties without this benefits can make this happen by applying reverse engeneering, then, franckly, I’ve lost the last bit of respect for them. If, OTOH
MS decides to use binary blobs as a competative weapon (like you have suggested), esp. given that there seem to be no valid technical reasons to go this route, there would imho be an even more compelling reason to get rid of binary blobs in the long run.
Still, this is all irrelevant for the discussion at hand. As long as the GPL only kicks in during distribution of derivative works, the installation of binary modules by the user is allowed. And as long as binary modules are needed by some users, developers will strive to eliminate the need for them. End of story
[1]http://www.codeweavers.com/products/cxoffice/
[2]http://www.codeweavers.com/products/cxoffice/requirements/
But I don’t know how tied current code is to the Windows kernel, but I do know that Outlook and Access do not exist in the Mac version. Access for sure has deep ties to Windows as does Outlook to a lesser degree.
Don’t kid yourself, it has nothing to do with code embedded in the kernel. If MS was using kernel hooks for Office, then Windows would be even more of a security concern than it is now.
Outlook and Access don’t exist for competitive reasons, they’re the two applications that most differentiate Office for Enterprise Users from Office for home/small business users. May as well throw Visio in that category as well.
MS created Entourage in place of Outlook in order to provide just enough messaging/scheduling for a complete office suite but lacking just enough Exchange compatibility to make it unviable for use as a substitute for Outlook in corporate environments. My gf was forced to ditch her corporate-issue powerbook for that very reason.
I’d suspect that MS could have the full Office suite running on linux within a year if they were properly motivated. The kernel licensing and binary blobs have nothing to do with it. They can DRM it, they can incorporate WGA, they can do a lot of things. The kernel doesn’t change anything.
I’d go out on a limb and bet that if desktop linux ever hit 10% penetration in enterprise markets, there would be an Office for Linux. Right now, and for very obvious reasons (yes, including low marketshare), it’s not in their best interests to.
NotParker, I don’t think you understand what a “binary blob” is in the context of this discussion.
Linus’ writing abilities are markedly improved since his first rants…
+1
It is becoming clearer why there is such a wide gap in hardware support in Linux. It is this attitude that demonizes hardware entrepreneurs just because they wish to protect their trade secrets.
Left to me Linus should stop all further Linux kernel development and spend one full year just working with hardware companies to get their PROPRIETARY drivers to work with Linux. Digital cameras, video, network cards, everything…… should just work when plugged in before we go back to the cutting edge stuff that will never impact 99% of the average users in Linuxland.
All we need is a dose of humility to go to these vendors and ask for their specs and sign no-disclosure agreements if necessary. If these GPL zealots then choose to fork the kernel, they are welcome.
I agree with Linus Torvalds that there shouldn’t be technical measures which keep me from watching a DVD or listening to music on whatever device we have available. Screw the silo mentality.
Fair use, a term which seems to be forgotten in our little globe of greedy bastards, should be the way to go.
He also mentions we should not try to assert Open Source copyright rules on other peoples code which wasn’t derived from Open Source coders. If they do take Open Source code, they should behave according to its rules.
If they `use` the code they should be `forced` to behave according to Open Source rules just because they `use` the code in a device.
I understand what he is saying. What’s the hang up? Simply, there is a difference between the copy and use of code in a device.
I guess it’s FreeBSD all the way, it’s amazing how champions of your rights can take them away just as quick as the bad guys.
it’s not going to happen. not that you are really interested, since you are just using this thread to bash Linux and GPL.
back to reality, this was not an argument between FSF and Linux. it was an argument between Linus and some of his developers. and as Linus and others in this thread stated, GPL is a licence that covers distributed code, so this idea of locking out end-users ability to install binary blobs was dead in the waters from the start.
Greg made an apology and also explained his reasons for considering such drastic measures, you really should read his side of the story before you start calling him a nazi, religious fanatic etc. while I don’t at all agree with the measures he proposed, I definately understand his frustration after reading this:
http://lkml.org/lkml/2006/12/14/63
of course, many of the posters here just saw this as an opportunity to bash their favourite hate subject, be it (Linux, GPL) and/or to promote their favourite OS (BSD, Windows). they have no interest in this story beyond that.
Actually, I like Linux, and as I have said in the past (and in this discussion), I use Ubuntu and before that Fedora Core. I did not know it was rejected until about half way through the comments, as the description of the article was not updated to reflect that. I have no problem with the GPL, I have a problem with people telling me what I can do and what I can’t with my computer, which is a common enough viewpoint on OSAlert.
I CHOSE the software I run, be it closed or open source, and I CHOSE my hardware. I have a real problem when people try to limit my choices due to their personal viewpoints. If they try to do that, then I will choose something else, it’s that simple. In this case, I would choose to not use Linux anymore, or recommend it to my clients.
And if you notice, I said I would chose FreeBSD, which is not actually my favorite OS, from a purely do everything I want standpoint, I prefer Windows, but I do not prefer MS, so I have tried other OS’s, and have settled on BSD and Linux, as they are more free. If this had of gone through, I would have dropped Linux in a flash, as it was a very hypocritical move, if you ask me
Right below the entry is an entry on userspace IO drivers ( http://www.kroah.com/log/2006/12/13/#uio ).
A large number of people have expressed interest recently in the userspace i/o driver core which allows userspace drivers to be written to handle some types of hardware.
….
This core allows for a very tiny kernel driver to be written to handle the interrupt generated by the hardware. Everything else can be done in userspace (direct memory access, interrupt processing, controller logic, etc.) In some instances, this framework has shown a noticable improvement over an all-in-kernel driver.
If I understand this correctly – For the nVidia and wireless drivers (which is what I assume everyone is so upset about), this would mean no more compiling or getting the exact distro/kernel/compilier/phases-of-the-moon package, and debugging kernels that previously had closed source bits in them should be a lot easier.
I think this is a good thing. Closed source drivers will loose that annoying quasi-legal status, binary compatibility will become less of an issue, and the entire system will be cleaner – which is a good thing.
Linus had a well written comment. To show things in the right perspective, this story should have been more about that well thought out long respond by Linus than about the short comment thrown by Andrew Morton.
If the Linux kernel would allow only GPL modules, it would probably mean that the kernel would be forked. And guess which version of the two alternatives most Linux users would be using if they had to choose whether to be able to use, say, the official nVidia supported high performance drivers for their expensive 3D/video hardware or not?
(edit: typos, typos, again…
)
Edited 2006-12-14 20:25
It looks like Linus put the kibosh on the fundies high hopes of GPL nirvana, but if it were to come true, desktop Linux would basically go away real quick. I’m sure at least FreeBSD would be happy to take up the slack thouugh.
The fact is, the reason I don’t think we should force the issue is very
simple: copyright law is simply _better_off_ when you honor the admittedly
gray issue of “derived work”. It’s gray. It’s not black-and-white. But
being gray is _good_. Putting artificial black-and-white technical
counter-measures is actually bad. It’s bad when the RIAA does it, it’s bad
when anybody else does it.
Enough said.
Well, we knew this would happen one day. GPL fundamentalism won’t be going into Linus’s tree. So what distro has the balls to go GPL only? Or rather which distro wants to go the way of the dodo bird for anything desktop related?
I’ve come back to this thread, after my tea, and maybe I’m getting old, but I’m a little lost. I’m out of points, and my rating through the floor, and justly so.
I really need a table with pro’s con’s on. A summery sheet. 200+ points is too much.
@BluenoseJake good work attempting to catch up there.
1st ReactOS, then Tux !
If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.
If anyone weasel words then it is Linus himself…
Whose opinion do you think the court will consider most when establishing what the license means? Maybe the person who actually wrote it….
Whose opinion do you think the court will consider most when establishing what the license means? Maybe the person who actually wrote it….
The courts do not take intent into account when adjudicating the meaning of a legal document.
And what little case law there is matches Linus’ interpretation.
Maybe he’s just got access to better IP lawyers than RMS had when he wrote the GPL.
The courts do not take intent into account when adjudicating the meaning of a legal document.
I hear this a lot – not sure I would agree but I guess it could be true.
Good thing the GPL doesn’t JUST use the word derivative huh? It goes on to set other criteria as well. The nvidia drivers certainly do not act independently now do they? If they were totally seperate then no “shim” would be needed huh?
Got any references about this set precedent?
So this means nvidia can get away with distributing a wrapper licensed under the GPL, and then asking you to separately download the binary blob to install yourself.
Remember, nvidia’s driver is not prelinked. It is compiled on your system, and the actual derivative is based created on your system. livna.org might be infringing, but probably not nvidia, and that is because they do distribute a derivative work. But trying to call nvidia’s binary blob a derivative is pretty specious
Grey area really.
if you are not happy stay with windows and pay
Considering he hasnt written any real kernel code in several years, is he really needed? He doesnt want to cooperate thats great, dont let him use your code in the Linux kernel, get it removed. Considering he owns a very small portion of code, that is pretty easily replaceable. Start a forked Kernel project and watch his crippled kernel die. Freedom lives in a kernel where Binary blobs are not allowed and the GPL 3 is the license. The Linux kernel doesnt belong to Linus Torvalds, it belongs to the community.
Edited 2006-12-15 02:15
Jeez, you make it sound like he’s only a “royal figurehead” and sits there twiddling his thumbs.. He hasn’t introduced new features recently, but he certainly still provides valuable code reviews and maintains the tree. Even so, saying that “he doesn’t write code” is not true because the act of merging other peoples code together may involve writing code as well as understanding the code you’re merging. (outside of simply/blindly applying a patch to the source of it’s parent’s diff)
You say:”Start a forked Kernel project and watch his crippled kernel die.”
I say: “Start a forked Kernel project and watch that crippled kernel die”
Freedom lives in a kernel where Binary blobs are not allowed and the GPL 3 is the license.
wow.. (shakes head in disbelief).. talk about getting religion. Your Stallmanistic idea of “freedom” is certainly different from mine. Like Linus mentioned, I don’t see how what you say is any different from DRM at all. It’s like comparing one theocracy vs another–the subject matter of their ideaologies are different BUT the nature of their effects on people are the same.
@jimveta
IMO it’s time Linus disassociates Linux with the GNU label and FSF.
FSF needs Linux. Linux doesn’t need FSF. Ditch’em.
Linux is a kernel. It’s already disassociated from GNU and the FSF.
The reason why people sometimes call Linux the OS “GNU/Linux” is that a lot of the basic userland utilities (not least of all gcc) are from the GNU project. And you know what? There’s nothing wrong with that.
I don’t think anti-Linux posters should concern themselves that much about internal kernel affairs…
@archiesteel
I don’t think anti-Linux posters should concern themselves that much about internal kernel affairs…
Agreed. But whoever stopped those who knew very little of windows to post comments?
No one…that’s because even Linux advocates (like me) usually know quite a lot about Windows, too.
Take me, for example. I’ve used Windows since the 3.0 days. I’ve installed Windows countless times. I still do “family and friends” tech support for Windows. Just last week I had to fix a friends’ PC who no longer booted into Windows except for Safe Mode, after he had caught a virus (and tried to clean it with McAffee).
Everyone is free to comment on anything, that’s the beauty of free speech. But when anti-Linux posters actively follow the LKML in order to capitalize on every argument there by turning them into imaginary schisms within the Linux community, it’s kind of hard not to see this as a bit pathetic.
@archiesteel
Everyone is free to comment on anything, that’s the beauty of free speech. But when anti-Linux posters actively follow the LKML in order to capitalize on every argument there by turning them into imaginary schisms within the Linux community, it’s kind of hard not to see this as a bit pathetic.
That’s each person in the Linux community duty to ignore and eventually the annoying ones will go away.
I’ve taken another means of action against the OSS fundies.
Linus is the only sane person left in the Linux world it seems. No political agenda not allowing users to chose what they want. Seems kind of like a dictatorship? I guess I’ll stay with proprietary systems that let me chose whether I run GPL drivers or closed source. I want a system that works, without other peoples fanatic beliefs telling me what to do. I stopped using Linux a couple of years ago and moved to OS X. No politics, simply software that works with minimal worry.
Linux on the desktop is going to suffer. There goes many mom and pop users who frankly don’t either a) know about freedom/open source or b) know, but it doesn’t make a difference. They will know longer be able to play Quake 3 with their ATi graphics card, or their wireless may no longer work without ndiswrapper (which even though its GPL, is makred as a tainted module).
Is nVidia going to release a pure GPL driver, and ATi for that matter? I doubt it. I highly doubt it.
Too bad, Linux has/had great potential, but GPL fanatics are beginning to get in the way with their own personal agendas.
When I do use Linux again, I’ll make sure it doesn’t restrict me with what drivers I use. Hell, not even OS X or… Windows does this.
Is nVidia going to release a pure GPL driver, and ATi for that matter? I doubt it. I highly doubt it.
It doesn’t have to be under the GPL. It could be under the BSD license (surely you like the BSD license?), the LGPL, or any other GPL-compatible license.
Can anyone name any other OS that is called the exact same name as the kernel it uses?
There is usually a userland system and a kernel, that AFAIK is the reason RMS says it should be called GNU+Linux and I can at least see the reasoning in it.
I believe that right now this is truly the beginning of the end of Linux, or at least GPL Linux. Maybe by some chance of luck most or all the GPL software projects (Gnome,KDE,X.org,alsa,etc……) will drop the GPL licenses and create a new free type licenses and move on under that.
Just an idea
Why would they? What good would that bring?
Oh, right, nothing.
There’s nothing wrong with GPLed software projects. It’s as good a license as the BSDL or other free licenses. The only people who don’t like it are software companies who want to take free stuff, tweak it, and then re-release it as non-free stuff.