Slashdot Mirror
Why the Novell / MS Deal Is Very Bad
jamienk writes "PJ from Groklaw has taken the time to really explain the big picture of the Novell/MS deal and how it all fits into the SCO case and the strategy some have employed to attack Free Software. If you thought PJ was becoming too shrill before, or if you haven't understood what the big deal is with Novell's agreement, it's really worth a read." From the article: "This is Groklaw's 2,838th article. We now have 10,545 members, who have worked very hard to disprove SCO's scurrilous claims, and we did. We succeeded, beyond my hopes when we started. But here's the sad part. As victory is in sight, Novell signs a patent agreement with Microsoft..."
I just popped over to google finance and saw that this had come in today, not mentioned in TFA: http://www.cbronline.com/article_news.asp?guid=2BF 9274C-A4EF-4A3F-8E14-ABFBA2178EF8
Can somebody who has been following this a bit closer explain this? It's getting quite hard to tell who is friend or foe any more...
And in any case, why bother... their stock is toast, so couldn't IBM just buy a controlling interest for $11.2M and wind it down?
... is against yourself.
Seriously though, I bet Suse loves all the good PR they're getting right now.
Could someone explain to me, in simple terms, how this effects anything I have anything to do with?
I use Ubuntu, why should this matter to me? If the Ubuntu folks don't like what Novell is doing can't they just ignore whatever Novell is doing?
Everyone is acting like this is the end of Linux as we know it. Honestly could someone explain why this is?
I don't know about anybody else but this hasn't induced any fear, uncertainty or doubt in me about Linux.
However, it has induced REAL fear, REAL uncertainty and REAL doubt in Novell SUSE. Up until this incident started, I had pretty much decided that SUSE would be the distribution we would base all our new web/db/mail servers on owing to its combination of corporate support and ease of use.
Now I'm back on the fence considering Red Hat or another distro.
Unfortunately, I think SUSE inadvertently screwed themselves. In this regard, I have to say that Red Hat is doing an awesome job. They have deliberately tried to meet the Linux "community standards" while still being commercial. If only they were more open with their non-Fedora distributions, we would have probably standardized on Red Hat from the start.
Sunny
Be my Friend
Everything I've read by Novell in responce to the criticism from the OS community tells me that this article means nothing. Had this article came out when the deal first happened it would be different but now that Novell has communicated that they didnt' want the IP clauses in it and they actually want to make MS and Linux easier to work together I don't believe this article. It just fuels the fud factories. Novell wanted to get MS to actually allow the Visual Studios and Office on a Linux platform. Stuff like that. This deal is to form a bridge between 2 companies that really should work together instead of figthing. Imagine a world where MS and Linux worked even 25% together than they do now. Here is an open letter to the community by Novell. And other releases.
That which does not kill me only postpones the inevitable.
Has Linus been heard from lately on this whole schomzle? He has resisted changing the kernel from GPL2 to GPL3. RMS and others (like PJ) are saying that GPL3 solves the problem going forward. Does Linus concur? Has he made any statement anywhere (like kernel.org) about how he sees the Novell/MS deal possibly changing his mind?
Just wondering
Look at the tomato! Isn't it sad? He can't dance! Poor tomato!
As I said, I don't have a problem with enthusiasm - others may feel different, because lots of peeps consider the display of emotions or enthusiasm as weakness, and they invest a lot of emotions in pretending to not care or feel something about these issues. But except for this last case, you will hardly find any criticism of groklaw and PJ that tries to address the issues she raises with logical arguments. I haven't followed the happening on groklaw in the past few years to closely, but since hell broke out about the Novell-MS deal, I have been regularly reading groklaw, and I have to say this: the oo.o fork article was not a trend, but an exception. PJ is still the same PJ I have come to know when the SCO case started: exceptionally thorough and logical in providing arguments to prove her point, but also a bit emotional and enthusiastic (which made some of her writings an easy target for those, who failing provide rational arguments concerning the points she made, dismissed her articles on the note of being "biased".)
Ok here is the deal as I see it through the eyes of a developer, sysadmin and ultimate decision maker when it comes to our linux environment. I am also the developer of quite a bit of GPL licensed code spread across the globe(not kernel). Novell and MS broke the spirit of the GPL with their little agreement, I don't care if it was legal, bottom line it is not in the community spirit. I will not ever run, recommend others run, help with others issues with any novell distributed software.
Bottom line I work for pointy haired bosses but they will not under any circumstances question what distro I run in the enterprise...novell put that in your pipe and smoke it. EV1 learned a valuable lesson and Novell should have taken note of what happens when you try to bend the rules.
... why bother... their stock is toast, so couldn't IBM just buy a controlling interest for $11.2M and wind it down?
Doing so would invite thousands of nuisance suits from people who want to be bought off by being bought out. Suing IBM would become both a neat way to make a few million bucks and an exit strategy for every failing company that could make a vuagely-plausible argument that IBM had something to do with their fall.
So instead IBM has chosen to counter-attack, sucking all the blood out of SCO and leaving a dessicated corpse hanging on a spike for all to see.
It's an object lesson on the pitfalls of trying extortion on Big Blue.
They have had this policy for a while. SCO is just the biggest band of fools-or-crooks to come along in a long time, trying something new with ramifications in one of the biggest-bucks fights in a long time: the war between Microsoft and Open Source. So SCO gets the biggest spotlight.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
At least Microsoft can't claim they're not involved in this anymore, like back when they were piping money through Baystar to fund this fiasco.
Weaselmancer
rediculous.
My inclination has always been to think of the freedom guys as a little strident, and a little too extreme. The things Linus says about licensing have always made the most sense to me intuitively, and the other guys have always come across as a little controlling, and a little crusading.
The one thing I've taken away from the Novell/MS deal, though, is that this stuff is really complicated, and it's really dangerous. I'll be honest -- I don't understand all of the implications of the deal, or why each of the two parties decided to do it. But I feel like something's going on -- like I'm playing 3 card monte on the street or something.
I don't think that non-specialists (ie., geeks who don't think much about law) are in a good position to know what's best.
Novell, and the guys that came to Novell when they bought Ximian and SUSE, have done an incredible amount of good for our community. We are, to a certain extent, depending on Novell's patents to protect us in this coming fight. I think they're good guys, doing what they feel they have to do in order to survive.
But even if this isn't nefarious, it's made us realize that we'd be open to something similar that was nefarious. Those crazy freedom guys weren't so crazy after all.
So I think we have to trust the people who understand these treacherous waters the best -- I think that's Eban Moglen. He says that GPL3 is necessary to counter this threat, and he says it will be effective, even if the kernel remains under GPL2. The toolchain will be enough to do what we need.
I don't want to demonize Novell, because they've given me a lot of great code, and because there are people there who are real heroes to our community. I think they're mistaken, and I think Linus is mistaken to stick with GPL2. It just ain't viral enough to keep us safe.
But instead of attacking people, or getting hysterical, I think the thing to do is to listen to our best legal minds, and back GPL3. So my feeling is that Linus's honor is beyond question, he's obviously a lot smarter than I am, and he might even be smarter than Eban Moglen. But when it comes to law, I'm going to listen to Moglen.
And I would say that the Ximian guys' honor is beyond question, and that they're a lot smarter than I am as well. But I'm still going to listen to Moglen about the law.
Again, my feeling is that we shouldn't let this break down cooperation, we shouldn't let it affect the civility of our community, and we shouldn't attribute bad motives to anyone. But we should play it safe, and innoculate with GPL3.
But I can't read the terms, because it's a secret. And they don't agree in public on what, exactly, the deal was.
Boy, am I relieved!
Help stamp out iliturcy.
That's why Shuttleworth was trying to poach developers. SuSe was funding key projects...many related to Microsoft compatibility... this agreement means that all those projects are now suspect for added code that shouldn't be there. The Samba team in particular was very proud of their implementation being very clean legally... the very soul of the team was sold right out from under them!!!
If such is the case we are really toast. A submarine-familymember-killer-spook is what we got here. I dont think there is any other way to get them but to pressure them into submission by a big PR move by the community NOW.
All enterprises making money off linux should say NOW if they support or not this kind of thing and Linus himself should make a statement (how about a special Novell cant use this clause for future releases of oo.o and the linux kernel? Yes, i dont care if debian rejects it or puts the friggin kernel in non-free).
NO SIG
Geez, I don't have the hardware drivers NOW.
:wq
one of "...Groklaw's 2,838[th] article[s]..."
Congratulations!
" ...
The phrase refers to the idea that a butterfly's wings might create tiny changes in the atmosphere that ultimately cause a tornado to appear (or, for that matter, prevent a tornado from appearing). The flapping wing represents a small change in the initial condition of the system, which causes a chain of events leading to large-scale phenomena. Had the butterfly not flapped its wings, the trajectory of the system might have been vastly different..."
http://en.wikipedia.org/wiki/Butterfly_effect
He asserts at length that it's bad, but I have no idea from this article what's in the alleged patent agreement, or how it changes anything.
Imagine that I have a Slackware system. (I do, even.) How is this system affected by any deal Novell signs?
I don't see how this can be connected, and since TFA doesn't describe in any real detail what the Novell agreement says, or how it has any effect on anyone other than Novell or Microsoft, there's not much point in just asserting that it's bad.
My blog: http://www.seebs.net/log/ --- My iPhone/iPad app: http://www.seebs.net/seebsfrac/
...on the Novell-MS deal. Seriously, PJ, give it a rest.
I wish I saw just as much criticism from you some time ago when the OSRM came out with that FUD on Linux infringing 283 patents. But noooo, not a peep from you at the time, since you were an OSRM employee.
I guess what Novell did wrong was not hiring you before signing the MS-Novell deal.
Humanity only wins when MS is killed off forever. Anything that helps MS in any way shape or form threatens our freedom and humanity.
-- Go ahead, burn this post. I was feeling a bit chilly anyways.
They don't sue Ubuntu because those guys are clearly NOT trying to circumvent the spirit of the license. It's Nvidia and their ilk that are doing that. And as long as they aren't actually distributing code, they can get away with it. Suing Ubuntu wouldn't hurt them directly, and would hurt some good people that are trying to do the right thing, so don't expect to see that happen.
What I do expect will happen, at some point, is someone with standing to sue will initiate a dialogue with them, and they'll remove the drivers. I don't think anyone is in a huge hurry about that, however, for the reasons outlined above.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Friends don't let friends enable ecmascript.
Well, OK, it says "The Novell deal was really bad, really really bad. It's good for Microsoft. What SCO is doing is good for Microsoft. The Novell deal is like the SCO deal because they are both good for Microsoft. The Novell deal is really stupid and Novell should know better."
And while I agree, nobody is going to be convinced by that. I would appreciate some deeper level of analysis and explanation. Perhaps something I could use when talking to people about it about why the deal is a bad idea.
Need a Python, C++, Unix, Linux develop
First, they did not do much to improve Yast2 on speed, bugs and the way it handles dependencies. This alone, made me avoid its Linux products.
Now comes this Microsoft deal. With this deal, I will not even touch its Linux products even with a 10 foot long pole!
Sadly, those who predicted that Novell will do nothing good for SuSE after buying it, (just like Word Perfect), are being proven right.
Linux supports more devices "out of the box" than any other operating system ever has. Yes, even FreeBSD.
r face.c
The other key highlight of this talk was:
Closed source Linux kernel modules are illegal.
Closed source Linux kernel modules are unworkable.
Closed source Linux kernel modules are unethical.
So who the hell is this guy? He's Greg Kroah-Hartman. Who the hell is that? He's a kernel developer. His name appears 149 times in my kernel sources (Ubuntu patched, 2.6.15). And, perhaps more tellingly it appears at the top of the files:
drivers/pci/pci-sysfs.c and
drivers/pci/search.c
both of which contain many functions which are called from functions in this file:
NVIDIA-Linux-x86-1.0-8776-pkg1/usr/src/nv/os-inte
What's that? It's wrapper for the closed source NVIDIA kernel module. What license is that under? The NVIDIA Software License. It's basically a proprietary EULA with a redistribution (without modification) exception for distros. It sure aint the GPL, or "as free" as the GPL (which is techically what the GPL requires for derived works).
So Greg.. why don't you sue them? You've made your position clear, fight them. If you havn't got the money, contact the FSF, assign your copyright to them, get them to fight. Given the choice between opening their source code or not being able to distribute their software at all, NVIDIA will choose to open their source code. How can I be so sure? Cause people buy their chipsets to integrate into things like set top boxes and other devices that run Linux. They need that embedded market, that's why they released the drivers in the first place. The problem is that no-one is making them choose.
How we know is more important than what we know.
When all you have is a conspiracy theory, everything starts looking like a conspiracy.
Conspiracists believe that all evil is attributable to one cause. Communist Insiders, The Elders of Zion, Vast Right Wing Conspiracy, Microsoft. But that's not reality. PJ needs to wake up and realize that SCO is able to be evil all on its own.
Don't blame me, I didn't vote for either of them!
For another fucking inane article on the deal we've all known about for 2 plus weeks!
Reading TFA, this came to mind.
So Novell has four large sales opportunities lost because the Microsoft IP FUD is working.
Re-read that, and put yourself in the seat of the guys who have a bunch of employees (with families and mortgages and kids going through college) and answer the question: continue to watch the company waste away? Or do something?
Hey - here is an idea: get Microsoft to admit that Linux matters, that interoperability matters, that customers who run Linux matter.
It's not like Microsoft is going to be able to keep pretending forever. Someone is going broker a deal first.
Should it be Red Hat? or Novell?
And while you are brokering a deal, get it in writing: Microsoft does hereby promise that their FUD argument does not apply to SuSE Linux.
It seems to me, that the authors of TFA want to have kept the status quo: Microsoft FUD to remain in place; Novell to continue to lose sales in the Fortune 500 because of that FUD, Novell's investment in Linux (and support of Linux and programmers) to fritter away because.... (I don't know. My guess is it comes from some deep-seated anti-establishment attitude.)
Point is: Novell had to take a bull by the horns - because so far, it was 0-4 in favor of the bull.
And all I hear is all this whining: but Novell didn't slay the bull! Novell should have forfeited the future! Novell took action, and didn't ask us first!
Like I said: nothing is impossible for the man that does not have to do it himself.
Regardless of whether this article has any substance or is merely a frothing rant on PJ's part, I think (inadvertently?) she's overestimating the adoption rate of GPLv3.
Linus has clearly stated that he intends to keep the kernel under v2, and most of the larger projects have yet to make any meaningful statement about it.
Never mind the scores of smaller projects that don't have the resources or prowess to make an informed descision about which GPL version to use; most of them will stick with v2. I wouldn't be surprised if most projects simply followed Linus' example.
So why is it every time Microsoft makes an announcement it's FUD, but this is supposedly legitimate news? FUD is FUD regardless of whom it's being used against, and this article is clearly FUD.
Embrace, Extend, Knife in Back....
DEAD DEAD DEAD DELETE ME
Donald 'Duck' Dunn: We had a band powerful enough to turn goat piss into gasoline.
The only reason programming gets any special consideration is because it makes pretty lights flash and calculations happen inside of black boxes.
Lawyers *are* a conspiracy. That's how they establish their credibility. They point to a mob of other unthinking lawyers and justify their claims that way. Just because you choose to ridicule and blindly insult someone without cause doesn't mean they have no cause.
Look at what PJ is saying again.
Novell is undermining the credibility of the GPL, and SCO is attacking the GPL in exactly the way you can foresee Microsoft possibly doing so. The difference is that SCO doesn't have a leg to stand on in their case. In a potential Microsoft case, they now have serious ammunition. Are you willing to risk everything that you're right and she's wrong?
Should we take advice from PJ because she's not a lawyer or because she's not a developer?
Tis very interesting that. Once again mentions Greg Koah-Hartman's talk. I just think it's a shame he didn't post the email.
How we know is more important than what we know.
Unfortunately, it wouldn't be a simple case to prosecute.
You see, Nvidia distributes two works they present as separate. A binary blob, and a 'shim' under the GPL, whose only purpose is to load their blob and link it into the kernel.
Also, Nvidia doesn't distribute the kernel.
So, while it's clearly illegal to distribute a working system using this two-part driver as part of the linux kernel, it's not quite clearly illegal to distribute just those two parts without the kernel, which is what Nvidia does.
They're exploiting a loophole in the GPL, and unfortunately, while I believe a court would probably rule against them in the end, once the entire situation was clearly explained, this would be an incredibly expensive proposition.
Suing distro makers that bundle the kernel, shim, and blob together in a usable form would be much easier, but you know what? No one really wants to sue a distro maker for trying to make their customers lives easier. That's a last resort, only if and when it becomes clear that simply talking to them won't do the trick. And that's how it should be. We're a community, and we don't and shouldn't jump to sue each other unless and until everything else has been tried and failed.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Friends don't let friends enable ecmascript.
A woman getting emotional. I'm shocked.
The shim is not under the GPL, it's under this license. Which permits nothing but redistribution in distros, with optional modification to the wrapper (or shim) but not the binary.
How we know is more important than what we know.
keep your friends close to you, and your enemies even closer.
Congratulations, now THAT's a good summary!!!
Isn't the real problem - and not just for Linux but in general - that it is MUCH easier to destroy, i.e. to be AGAINST something, instead of FOR something and to DO something? So of course, lazy as people are (for evolutionary reasons, food was scarce until only *very* recently! (yes yes, and for many even today)), they all prefer to be against something rather than do soemthing they are for. Important words are "be" vs. "do".
I've a suggestion for all those MS/Novell haters (but by the way, this very much reminds me of the communists in Germany in 1933 who fought the left - but not left enough for the communists - party "SPD" a lot harder than the real enemy of both of them, fashism. I guess it's human to go with much more religious fervor against your own people rather than the enemy if one feels they stray from the "one true path"). DO something! Sue someone. Organize a march on the MS HQ. Hand out leaflets in your hometown. Somehow the one thing I truly don't like is everyone talking, and worse: imagining things that just MIGHT happen! It's Iraq all over again (yup, maybe unfair to use this comparison, unfair but maybe not wrong - in both cases the spin doctors played on the public's fears and used intimidation and speaking loudly whenever a voice of reason attempted to be heard) - lots of fear of what MIGHT happen! Most postings I've seen against Novell only ever refer to secrets, might-be's, *perceived* threats (or does anyone have any PROOF - and I would count a GOOD argument, I don't even ask for hard numbers, that the threat of litigation to anyone has increased (instead of remaining the same or even decreasing)?)
Well, if kernel developers are pussies, than who the hell needs them? We'll just go back to waiting for HURD. After all, their manliness is all that really matters.
Why would you want to sue people for writing their own code and releasing it under whatever license they choose? That would be like Microsoft going after people writing GPLed software with Visual Studio - unethical and stupid.
Cause, simply, the license on the Linux kernel says that if you write modules for it, those modules must be licensed under the GPL or you can't distribute them. If you don't like it, don't use the Linux kernel.
:)
Oh, and your analogy, stupid as always, thanks for holding up the Slashdot tradition
How we know is more important than what we know.
Novell doesn't own the open source work/code, Microsoft doesn't own the open source work/code.
What is sooo wrong in the open source license/structure/community whatever, that a deal betweem these two entities or any entities whatsover, can be a threat to open source?
It doesn't make sense
I don't have to follow other people's license for writing my own code. Compaq clean room re-implementation of IBMs BIOS is a case in point. Mere APIs can not be copyrighted and its a good thing, otherwise we surely wouldn't see Wine. Think of it as another example of fair use.
SCO
a) Get 5% s a finders fee paid back
b) New contracts become SCO's in its' entirety
Also note that Novell bought the complete rights for a billion. SCO paid a score of millions. What did SCO think it was getting for such a small ammount? The whole she-bang?
Well, for one thing, Novell owns a company (now a division) called Ximian. The people behind Ximian are the people who originally developed GNOME. I believe that they are still active in the Gnome project. If I were you, I'd think about switching to Kubuntu.
Perhaps you are referring in part to Miguel de Icaza, who was among the founders of both GNOME (in 1997) and Ximian (in 1999), and is very positive about the Novell-Microsoft deal (which got him a lot of ill-will recently), but naturally other Ximian people would have been active in GNOME, since Ximian was in the business of GNOME-related things. Still, to leave GNOME because of this makes as much sense as leaving all other OSS projects that Novell employees contribute to, OpenOffice among them. I'm not saying it doesn't make sense, mind you, just that if it does make sense, then you need to stop using a lot more than GNOME.
Personally, my approach to this is to see what GNOME does with the GPL3 license. As the official desktop of the GNU project, I would be surprised if GNOME doesn't move to the GPL3. Assuming it does, most of my worries will be alleviated. In particular, if one of GNOME/GTK+ and KDE/Qt move to the GPL3, and the other does not, I will in all likelihood move to the one that does, whichever it is.
Until today, I wasn't sure about GPL v2 vs. v3 - now I am. v3 is definitely the way to go. Sorry, Linus, but you're wrong. You are a software engineer, not a lawyer.
IANAL, but I play with enough of them in my day job to spot it when they've smelled blood. And right now, GPL v2 lies bleeding in the water with the sharks circling it.
Assorted stuff I do sometimes: Lemuria.org
A much more plausible "butterfly effect" is found in Bradbury's 1950's story: http://en.wikipedia.org/wiki/A_Sound_of_Thunder
Novell, IBM et al. are not Linux aficianados because they give a damn about the community, they are doing it because they make money from Linux, and because it indirectly hurts their main competitor, Microsoft. So any notion that these companies are good guys beyond their own self interest is misplaced. It just so happens that their interests and the interests of the community, at present, more or less concur. But the Novell/MS deal shows how fragile that consensus is.
Microsoft and Windows aren't going to go away overnight. And, ironically, the more Linux is adopted by companies who currently use MS, the greater the need for interoperability. My guess is that Novell thought it was gaining an edge in that market. They thought increased interoperability would lower the threshold for companies adopting Linux, because they wouln't have to replace all MS systems overnight, as long as Linux and MS systems could work together. And with Suse having unique access to interoperability API's, they would be the Distro of choice when it came to heterogenous solutions. The problem is Novell didn't anticipate the fallout from the deal. They didn't anticipate Ballmer's post-coital comments. They should have.
Microsoft is just the last company jumping on the Linux hype, as you can read in my blog here: http://www.thinsia.com/blog/ The war between the virtual machines has begun!
Then suppose you don't make code that patches into the source files but you do distribute a whole extra source file to be included at link time. Perhaps it provides its own implementation of printf() to be used instead of the C library's one, or something. Is the copyright situation different from if you'd made a patch to the source? Again I can't rule for certain, but I think that in most jurisdictions your extra object file would still be considered a derived work of GNU grep, or at least, you couldn't distribute copies of grep with your extra file already included. Perhaps it would be different if you just distributed the
-- Ed Avis ed@membled.com
... patents. We need to get ride of those stupid laws.
wtf.n0x.org
Do I understand this properly?
MS signs patent agreement with Novel.
Novel introduce MS patent based technologies (WMV, DRM, interoperability stuff, etc) into their variant of Linux (while also keeping up to date with the kernel release, etc, done by the community).
People deciding which Linux to use have a choice between the souped-up version from Novel which will do lots of Windows stuff or regular Linux.
Regular Linux can't take the MS stuff into itself because it's patented. The MS Linux can take all the regular Linux stuff because it's GPL.
Result? MS start to get significant leverage over Linux by proxy, by trying to dominate the Linux variant market with "their" (e.g. Novell's) variant. MS would do this itself I think, without going through Novell, if they had the linux know-how.
MS then start putting seriously nice stuff into MS Linux, more and more people start to use it, become dependent on it, MS start making it easy to migrate from MS Linux to Windows, etc, etc - all sorts of strategies are available at that point.
I can see why a modified GPL is needed now, to exclude patented material from the kernel; this is to protect Linux from MS's embrace/extend/exterminate policy.
WTF are you talking about? I just looked at the Linux kernel license, and it's GPL. The GPL covers the kernel source itself, not third-party kernel modules. This means, that if you were to write a driver for Linux that is not included in the kernel source, your driver would not be part of the kernel, but would be your own work that you could put under the helm of an arbitrary license. This is exactly what NVIDIA is doing.
NVIDIA just wants to protect its right to conceal its chip architecture from its competitors. After all, NVIDIA DOES provide drivers for Linux, Solaris and FreeBSD. And if they were to pull them, users could still migrate to different hardware.
(And besides, which 3D accelerated graphics adapter would you suggest that has a GPL-ed driver?)
A broader base of drivers that can (and will) be shipped with GNU/Linux distributions only helps to enlarge the user base.
Nobody wants to use Windows. Me neither. But I definitely wouldn't run Linux if I couldn't make use of my hardware. So, NVIDIA's driver is the ONLY reason I'm using Linux. I have an NVIDIA card. I have no interest in using Linux with the non-3D open-source NVIDIA driver that comes with X.org (that'd be like using standard VGA in the year 2006). And I'm sure many people feel this way about the issue.
This is a very good question. But consider: if you wrote a few extra lines of source code to add some feature to GNU grep, for example, and then you distributed your extra code as proprietary, so you have to pay $10 for the privilege of patching it into grep and rebuilding... you can think about the ethical issues, but more importantly in this discussion, what is the legal situation? Is your bit of extra code a derived work of the grep source code?
Thats actually a really pertinent example.
For instance, suppose Novell did something like this so that grep could pattern-match in MS Word documents... something thats not altogether too far-fetched to consider.
In the free world the media isn't government run; the government is media run.
Do the posts here really matter?
It is the CIOs, CFOs at companies who decide what to buy. If Novell or Oracle offers them a better deal, they would go for it.
If Novell tells them that they have an agreement to make Suse work better with Windows, they would view it as a point in Novell's favor (regardless of what others think).
Not Bill, Uncle Fester. They didn't do the Novell deal because they cared nothing about Linux and its supporters.
Gerry
"What is sooo wrong in the open source license/structure/community whatever, that a deal betweem these two entities or any entities whatsover, can be a threat to open source? It doesn't make sense"
It sure don't make sense. What is wrong about is that the ink was hardly dry when Ballmer and Hilf was using the agreement to threaten other Linux developers with litigation unless they signed a similar agreement. Finally, as a Novell customer I give Novell no authority to negociate on my behalf with Microsoft. Novell should cancel the agreement immidiatly and return to the fold. Does this answer your question.
was Re:The Big Picture
davecb5620@gmail.com
- It's way more fun to stomp SCO into the ground.
and dance around the smoking crater.
- Denying the FLOSS community the satisfaction of (1) would alienate people
- SCO's stock is likely going down from where it is.
- Buying SCO would invite other nuisance suits against IBM
- SCO has poison-pill provisions.
- Most of the heavy lifting is already done. The rest of the suit could quite possibly cost IBM much less than 10M.
- IBM could recover money from their counter-suits (From SCO itself, and possibly also executives, directors and even the lawyers).
- SCO is now very unlikely to get any money from IBM.
So, yes IBM could buy SCO with their petty cash reserve, but it would be a silly thing to do.Sometimes boldness is in fashion. Sometimes only the brave will be bold.
Nothing, except that the code that neither of them own is the code covered by their agreements? How can two companies make an agreement over code they don't own?
Even if that doesn't show how silly you are, think on this: the code ownership is covered by copyright. The agreement is to do with patents. It trumps copyright in as much as you can write the code, but as soon as you run it, you need a patent license.
And what use is code you can't run?
Microsoft stole everything they ever did. What about the Zune? Infringing on anything there, buddy?!
I guess it's a good job for M$ that ALL black hats use Winblows, at least this way, you won't be pissing off anyone you REALLY shouldn't!
I just looked at the Linux kernel license, and it's GPL. The GPL covers the kernel source itself, not third-party kernel modules.
kernel modules are derived works of the kernel. As such, they are required to be GPL licensed. This is pretty fundamental stuff to this very complex conversation the rest of us are having.
The debate is over whether or not NVIDIA:
1. are distributing a derived work
2. is bound by the license on the kernel if they are not distributing the kernel
3. is satisfying the terms of the GPL by distributing partial source code to their driver
4. would only be violating copyright if they were distributing binaries with no source code wrapper
5. are guilty of contributory copyright infringement by facilitating others to distribute binaries if 4 were the case
We're all sipping martinis and talking about this stuff, you're still trying to find your shoes and wondering what tie to wear. Come join us when you catch up.
How we know is more important than what we know.
For your own use, you do whatever you want. Distributing your code based on GPL is entirely different.
Patents Drive Free Software as Hurricanes Drive Construction Industry
"PJ sometimes tends to get emotional - which is absolutely fine by me .."
Shouldn't these comments be directed at the relevent web site. Apart from her emotional state what do you have to say regarding the NOVL/MS deal. I notice in your long response you managed to not mention any of the points in the original article by PJ.
"Groklaw is no longer a reliable source of information. Because it is so obviously biased"
You mean like Hilf or Ballmer are unbiased don't you. Will a non Novell version of OpenOffice.org that includes support for OpenXML be in violation of a MS patent.
was Sometimes (Score:4, ad hominem )
davecb5620@gmail.com
I am TheRaven on Soylent News
The last distribution that made a shady deal with MS to attack Linux was Caldera. Would you deploy Caldera today?
Where is that guy who'd die defending what I had to say when I need him?
It seems like this is an efficient way to make sure that the code is clean.
Why would Ubuntu want to distribute those proprietary drivers with their distro? Isn't the whole idea that you can add drivers on your own? If the driver is proprietary, and the OS is open source, does that stop me from using the driver if I buy that video card? And if I don't use that video card, why should I have to have all that extra stuff in my distro?
Please excuse me. I don't follow the minutiae of the Open Source community as closely as I'd like. It's all so complicated (and math is hard).
On a side note, it surprises me that no commercial outfit has seen fit to put up any challenge to the Operating System Wars. OS/2 has been gone a long time. Why has no one else stepped up?
You are welcome on my lawn.
Dude, have you actually heard what Steve Ballmer has to say? He cares less about making money and more about destroying the competition.
Please, for the good of Humanity, vote Obama.
Er, dude, no... Novell's work on SuSE is all released as free software, and that includes the plugin they're writing for OpenOffice to make it compatible with Microsoft's file format.
-- Ed Avis ed@membled.com
I don't get it?
;-p
I mean, isn't Windows a free OS, too?
I mean, just now I downloaded a copy from my local Bitorrent store and it didn't cost me a cent.
Personally I think it's all a tempest in a tea-pot. Ten years ago something like this may have mattered. Nowadays there are to many large companies with vested interests in Linux, Open Office dot Org, and other free software projects. Many of these companies have large software patent portfolios and some are willing to go to launch an a retalitory strike should Microsoft start suing Linux users or distributors over patent infringement.
Furthurmore, the solution is simple. People who are pissed at this deal should spend less time boycotting Novell and more time badgering their local legislators to invalidate all software patents. If I am wrong and the MS/Novell deal actually does matter, it will come to nought if software patents become unenforceable.
Do I detect a sexist overtone there?
The hole question is not that they have a "loader" that links the binary blob to the kernel. What they say, and Linus also agree, is that the binary blob is not based on the kernel, it is based on the windows driver and their hardware. since the binary blob is not a derived work, only the thin linker layer is, then the source for this part is not bound to the GPL.
How to prove that this binary blob has not taken any part of the kernel when it was made is a tough job to me, and that is why no one has came out and sued nvidia, yet. Besides who has the money to contract the top-of-the-line lawyer that this would need?
[]'s Victor Bogado da Silva Lins
^[:wq
How we know is more important than what we know.
GNU/Solaris especially if, as rumored, Sun put Solaris under the GPL
Donald 'Duck' Dunn: We had a band powerful enough to turn goat piss into gasoline.
Read the Halloween Memos. They do, in fact, expend effort to crush what they see as their only real competition: http://www.catb.org/~esr/halloween/
I use Gentoo so this shouldn't matter to me.
When I'm finally done compiling the whole system, this problem will be already over.
The problem with the Microsoft deal has to do with the ramifications of Novell actually paying licensing to Microsoft for any and all Microsoft patents. This is one of the heavy hitters in Linux, who arguably owns the right to a lot of Unix IP signing a licensing agreement for a bunch of patents: if you sign the damn license you can never litigate the patent's validity. Given the state of the patent office, this does not bode well: Novell has acknowledged the validity of all present and future Microsoft IP by this agreement!
IAAL, but I don't have time to run down all the implications of this, but it looks a lot like Novell has sold out to the evil empire.
Dude, have you actually heard what Steve Ballmer has to say? He cares less about making money and more about destroying the competition.
That's easy to say when you're worth billions. Don't kid yourself--it's about money. By destroying the competition, they can make more money.I saw it on Slashdot, it must be true!
...you could say that MS and OSS are locked in a struggle, eh?
Blar.
1. Novell agrees to violate the clear intent and spirit of the GPL in an attempt to comply literally with the words but not with the actual known purpose of the license to make money off of code Novell didn't write and doesn't own. So instead of trying to prove the GPL isn't binding, they just kick it to the curb and step over it and dare the community to do something about it?
2. puts a FUD legal cloud over Linux (this time a patent cloud) or in any case an "IP" cloud, as per Steve Ballmer's vague wording -- and was Darl McBride's less vague?
Actually, this makes GPL software from Novell safer, not less safe, for others to use. Why? Because Novell cannot distribute software under the (L)GPL if that software is covered by patents. Since Novell has negotiated a legal agreement with Microsoft, they can be presumed to know what patents apply, and Novell has to be extra careful about patent violations when distributing software, since they would be liable for willful commercial infringement.
3. makes Novell's Linux cost more, because it has agreed to pay Microsoft royalties, whereas SCO asked for money for its license;
Novell made a huge net profit on the deal, so Novell Linux won't cost more.
And since Microsoft paid Novell and not the other way around, the logical conclusion is that Microsoft does not have IP of any value related to Linux or Mono, and instead is using this as a marketing gimmick. And courts probably won't see that any different either.
I see a dusty little tombstone with a penguin logo, tucked away in an inaccessible corner of Microsoft's new Graveyard.NET
They don't call it "embrace and exterminate" for nothing.
Don't let THEM immanentize the Eschaton!
The more money you have, the less money is worth to you.
As controversial as this is going to be I should add this bit of truth. You can badger congress to outlaw software patents and they can say OK lets outlaw software patents and then attempt to draft legislation to do so. The patent professionals will voice that fact that software was at one time non statutory patentable subject matter but the situation was unworkable due to the fact that systems were worded as partly hardware, logic could be embedded in hw or remain as software, etc. There was no bright line rule between software and physical systems. Many attempts were made to distinguish between the two but they were all circumventable. This is not mere postulation of what will happen if software patents are outlawed but is the reality of what did happen. The real solution lies in the courts raising the bar on obviousness standards to the execeptional contribution standard and then 99 percent of software patents would be deemed invalid due to the unexceptional amount of innovation contained within. Perhaps the other one percent should justifiably be patentable so that true innovation would be encouraged.
The whole "making money" thing only addresses the motivation for DESTROYING other companies. It doesn't negate the end result. OSS just happens to fall into the category of "other companies" from the point of view of any corporation like Microsoft. Microsoft wants to destroy Linux collectively just like they would be interested in destroying Boland or Netscape or Oracle.
A Pirate and a Puritan look the same on a balance sheet.
Ok, so nVidia should GPL the shim, but keep the binary blob as it is.
Then everything will be ok.
Software that uses an API does not become automatic derivative works of the API provider.
How would proprietary user-space software work otherwise?
Just because some software uses printf() does not mean the software should be GPL because the software runs on Linux (which has GNU libc).
Especially since printf() is implemented on many systems (and libc's) which have alternate licenses.
I'm sure Microsoft Windows has printf(), and I'm sure it's not GPL.
- 51acf00e7b6e2977807a371f24007eb39a863e7f
Correct, but they are not 3D accelerated graphics drivers. Just graphics drivers.
Do you really think they want nVidia to take their ball and go home? Do you really think they want a linux kernel with no 3D hardware accelerated graphics drivers? Because if you sue ATI and nVidia and they play hardball, that's what you will have... I think they let it slide because they prefer to be competitive with Windows.
I think he cares about destroying the competition as a means to make money. MS's business strategy can be summed up as "Establish a monopoly, and the money will follow!"
A long time ago, Gates said the most important thing was to control the standard. This has always been msft's focus. And it has worked like all hell.
If ODF caught on, msft might lose control of the document standard. Nearly msft's worst nightmare. Consider msft's aggressive lobbying in MA.
Msft is giving corporations yet one more reason not to use Linux. Msft is saying that Linux is a legal mine-field. This is the reason that msft funded the scox scam. Now that the scox scam is winding down, msft needs a new company to keep the scam working.
JMHO.
If there is something wrong with PJ's arguement, please point it out.
No point is just giving your opinion about PJ.
This is not fair use, it's how the copyright law works. In regard to computer code it specifically allows you to produce code that does the exact same thing, as long as you take the "clean start" approach. In simpler words, you have to work for it, not simply copy&paste.
Fair use is another section of the copyright law, which says you can use copyrighted material for free and without requiring permission in certain cases: teaching, news, criticism, review, research etc.
(Just putting things in place. A lot of people confuse fair use with "regular" copyright.)
i ate crayons when i was a kid and now i have two braincells and the blue ones taste nicer
Why is there so much confusion here? Does no one read the GPL? You don't have to be a lawyer to understand that GNU libc is released under the LGPL, and if it were released under the GPL instead then any program using printf in glibc would be required to be GPL as well.
--jeffk++
ipv6 is my vpn
Is there something in the article you specifically dissagree with? If so what?
As I understand, msft agreed not to sue novell customers for 5 years.
After 5 years: then what?
1) What if you are using redhat?
2) Novell can already distribute 100% legal GPL code, can't they?
3) Why are the terms of the agreement secret?
4) Why should novell be given exclusive rights to a public resource?
which 3D accelerated graphics adapter would you suggest that has a GPL-ed driver?
I suggeset Intel, since you asked...
If you're a zombie and you know it, bite your friend!
I'm pretty sure I could legally write a plugin system for grep which used binary plugin modules, release the plugin system's source, and then release closed source plugins. This seems reasonably comparable to what nvidia has done.
If IBM were to settle with SCO, it would leave too many legal questions open. That leaves the field open for another IP leech to use the same tactics as SCO. If those tactics are launched against a startup company or small organization, they won't have the resources to fight such a protracted court case. As long as cases settle out of court, the law continues to be a cudgel for stifling innovation and competition, when the IP laws were intended to protect and encourage those aspects of business.
IBM generates a huge portion of their revenue from OSS through their consulting services. To continue the battle is the best of both worlds: they get paid for providing an honourable defense of OSS.
This is the first time I've disagreed with PJ on an issue. The Novell-Microsoft deal does not affect GPLv2 or v3 software, because the clauses are written such that any IP attack invalidates the GPL for the jurisdiction in question. A side-deal does not miraculously create a new license that would grant Novell continued access to the disputed code -- they'd be forced to cease distribution along with any other US-based vendor (presuming the case were launched in the US.)
Both versions of the GPL are very clear on that.
As to side-line technologies like Novell implementing software that can access Microsoft's document formats, the issue may be a bit fuzzier. You can't branch GPL code and integrate it to proprietary software that will be used or accessible outside your own organization, regardless of licenses to the proprietary components. Any such modules would have to be isolated similar to NVidia's video drivers or the code for Oracle, Sybase, and DB/2.
In the event that OSS zealots decide to take up the battle against any and all proprietary software running alonside or on top of a GPL-based kernel, OSS will die for any use other than hobby programming. Even video games would no longer be ported to OSS platforms, video drivers would be yanked from distribution, and the OSS zealots would be left with nowhere and nothing to run.
Zealotry does not buy any benefits except stroking the egos of the zealots themselves.
I do not fail; I succeed at finding out what does not work.
Ironically, Nvidia can not likely open source their drivers for the same reason that ATI cannot ... they likely infringe each other's IP.
Do users deserve to get full functionality today on the most powerful and popular hardware components in the market or should they be locked out until enough underlying issues are resolved that the drivers can be opened?
Incorrect. All DRI drivers are 3D accelerated. My Intel graphics hardware works very nicely in FreeBSD with the drivers Eric Anholt and Keith Packard wrote. Both of them now work for Intel, and continue to write graphics drivers under the same license as before, only now they don't spend so much time supporting ATi hardware.
I am TheRaven on Soylent News
I'll try not to rant, but here's a couple of things to everyone should consider... The referenced keynote speech is truly detailing more personal opinions than citing facts... The references keynote speech is not entirely accurate, the SLED 10 does indeed provide access to proprietary drivers and proprietary applications through Novell hosted and/or sponsored repositories. Greg is a very outspoken person in regards to his open source beliefs, you can verify that by doing a Google search for his name - some might even argue he reaches the zealot level... Most -important-, I don't believe the "viral" concept of the GPL has been tested in any major court of law to this date. I am referring to the concept of "derivative work" in the language of the GPL license being applied specifically to things like third party kernel drivers. I don't believe there is much disagreement and I do believe the courts have upheld that you cannot take a piece of code released under GPL and not comply with it's licensing agreements, but I do not believe they have ruled on what the licensing obligations are to -interface- with a piece of GPL code and whether or not the license could become "viral" in nature. Please, if you do have a -citing- of a legal case that deals with this, then post it. The impacts of "viral" licensing far exceed the limited focus of GPL licensed software.
We should look at the PR aspect of this: wouldn't suing Nvidia make other hardware manufactures more reluctant to support Linux? And what about the strategic aspect? Isn't 3D critical to winning the desktop (not to mention niche fields like scientific visualization)? Given that NVidia is the primary option for 3D acceleration on Linux, wouldn't it be a little short-sited to spend a few years in court getting the source code to today's graphics cards when it'll mean that we can't use tommorrow's hardware?
-1, Too Many Layers Of Abstraction
Interesting that you mention Sun.
.... Well I push for Redhat Linux while others favor Windows. It goes one way or the other. Either Redhat will get our dollars or Windows. I don't see us increasing our IT budget and buying more of both. In at least this little microcosm it is a zero-sum game.
Were I work we are in the processing of moving from Sun to
America would be much better if people stopped believing everything that "someone says". Just because Richard Stallman says my code is derivative work doesn't make it so any more than making brief references to Harry Potter in my own book makes it property of J.Rowing. People have been saying lots of things - "I didn't have sex with this woman", "There are no american soldiers in Iraq" and "more doctor's smoke Malboros". If a court case came up where I was asked to turn over my source code just because I was calling someone's APIs, I would win based on fair use, or unfair contract or something.
It's also an object lesson on the pitfalls of lying one's ass off about everything. Sure, IBM had the lawyers, but SCO's case never had merit to begin with.
- First they ignore you, then they laugh at you, then ???, then profit.
Now I just wish Intel would make some higher-end discrete graphics chips.
"[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz
Nothing novell has done legitimizes any MS lawsuit. It might be perceived that way but in the end the agreement is only that MS and Novell won't sue each others customers - thats it.
> And one of the most compelling features of the GPLv2 is that you didn't have to be a lawyer to understand it, and it left very little "wriggle room" for dispute. Its not simply that it stands up in court, its that it is so clear that it usually doesn't get that far. It should not be lightly dumped
The problem is that outside forces are now trying to control us via hardware, as well as software. The GPL v2 protects vs. most software or contract based threats, but the GPL v3 will stop them from restricting the Four Freedoms via hardware or contracts as well.
The TiVo itself isn't that bad per se, but when that starts happening to computers in general (e.g. they're restricted to Windows-only, or to any specific set of programs chosen by the hardware vendor), we're screwed.
As for the Novell deal, the problem is that it makes two classes of Linux users: Novell customers (who are immune to certain lawsuits), and non-customers who are not. It also gives Microsoft a place for patent FUD ("This agreement proves that there are IP issues with Linux." type of stuff).
More troubling is what other inventive contracts they could come up with--how they could use them to fracture the Linux community and divide us legally. Like the old saying goes "we must hang together or we will surely hang separately."
I'll give Novell credit that I don't think they had bad intentions, but I do think they're being used like a $2 whore, and that sickens me. I hope they find some way to come clean soon, and that the GPL v3 will protect us from any further shenanigans.
Unfortunately, as you may realize, while simple and clear contracts are wonderful, lawyers aren't so nice. Just like you sometimes need evil, complex code to perform some difficult and nearly impossible computation, sometimes you need an arcane legal contract to protect you from changes in the law and inventive legal attacks the other side's lawyers will come up with.
So, just as I don't trust management to understand technical matters, I don't trust coders (including myself) to understand legal matters. IANAL, so I'll listen to Eben Moglen, because he's the only person I know that we can trust to protect us from legal threats we probably don't even comprehend. Sure, it might be fun to play lawyer on Slashdot and opine about how the law should be, but it's almost never cut and dried--many laws contradict each other outright, and it takes far more legal skill than I have to figure out what does or is likely to take precedence in a court of law.
off topic, but can you recommend a distro over suse? I switched to it because it was the best KDE distro I could find... but yeh, I seriously hate yast, and zen has not been smooth for me. I loved Ubuntu for the most part (except... gnome)... but had a deep hate for kubuntu. Although now that I know more, i'd be willing to give kubuntu a shot again and make it a decent KDE distro. but would you recommend anthing else? Also, what do you think of Mepis over Kubuntu? or PCLinuxOS? or gentoo? I realise this is off topic, but you outlined the problems in suse that piss me off, so i figured i'd see if you had any advice.
writing in ungrammatical english but not uninformed, Russian lawyer Czarkoff says in OSNEWS that indeed this deal signals that linux sysems dev is selling out to the dark side. Once it looses its soul, its marketshare won't matter.
SLASHDOT: news for people who can't concentrate on work or have no life at all and got tired of yelling back at the TV.
There is substantial reason to believe that kernel modules are not derived works of the kernel if they use the standard published kernel APIs, just as normal userland programs are not considered derivative works of the standard system header files and libraries, even if the userland program #include's or links against libc.so.
It's also true that the Linux kernel includes a fair amount of BSD-licensed source code, so it's obviously not true that all Linux kernel modules need to be under the GPL to be redistributed. A quick check against the Linux-2.6.0 kernel suggests 274 examples:
% find linux-2.6.0 -print0 | xargs -0 grep -l 'Redistribution and use in source and binary forms, with or without' | wc -l
274
These files start with linux-2.6.0/arch/m68k/mac/iop.c, include the ACPI & scsi/aic7xxx drivers, NFSd (linux-2.6.0/fs/nfs/idmap.c) & SunRPC (linux-2.6.0/net/sunrpc/auth_gss/auth_gss.c), and others.
1. are distributing a derived work
While Eben Moglen may have other opinions, most people on the OSI mailing lists did not consider the nVidia driver to be a derived work of the Linux kernel.
2. is bound by the license on the kernel if they are not distributing the kernel
The answer to this is obviously no.
3. is satisfying the terms of the GPL by distributing partial source code to their driver
The answer to this is obviously no, but the nVidia driver isn't and never has been under the terms of the GPL, because nVidia wrote it themselves.
4. would only be violating copyright if they were distributing binaries with no source code wrapper
The answer to this is "probably no", going by the "published API"s standard I mentioned above for deciding whether source code constitutes a derivative work or not.
5. are guilty of contributory copyright infringement by facilitating others to distribute binaries if 4 were the case
This suggestion comes straight from Eben Moglen, and is unlikely to gain much traction in a court of law. (a) nVidia has done nothing to encourage people to redistribute a Linux kernel plus the proprietary nVidia driver. (b) nVidia releasing a free (as in "no cost to download") driver for someone to use their nVidia video card with Linux would almost certainly qualify as "fair use" and be protected under 17 USC 107.
And if not, just who would you sue? Linux users who want GLX hardware acceleration?
SCO sued it's userbase, and look how far that got them...
"The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
2) Novell can already distribute 100% legal GPL code, can't they?
No, they can't. Novell can't distribute code under the GPL if they know that it violates patents.
1) What if you are using redhat?
Novell's protection is meaningless because if there is actually a patent that some GPL'ed project violates, then nobody can use the GPL'ed software in question, not RedHat customers and not Novell's customers.
4) Why should novell be given exclusive rights to a public resource?
They aren't; their agreement with Microsoft is meaningless.
Groklaw is basically an attempt by a fundamentally autistic community to create some kind of understanding/interface into a neurotypically created legal system. Good luck with that, guys. What you don't seem to get is that the single biggest difference between the neurotypical population and you is that no matter how vitriolic or abusive you might be of someone, you come out and do it openly. A lot of autistic people I've known barely have the neurological capacity for lying...and lies are what the political and legal system both are built on.
I've also had it up to here with the paranoia and persecution complex being displayed by the FSF and its' various fanboys. They see deadly enemies under every rock and behind every curtain...and it's getting very, very old and tired. The other thing is that the definition of an enemy in the FSF's worldview is literally anyone who doesn't worship RMS as God and follow their party line to every microscopic degree...which basically means around 98% of the population.
In case any of Stallman's zombie army who usually posts on here feels like answering this and telling me that I'm not paranoid enough...Try realising for a change that it's *you* who are fucked in the head, not me. The world at large is not going to accept the kind of totalitarianism you believe Microsoft are trying to foist on people...I am NOT going to believe that the antritrust trial happened solely because of Stallman and his fellow Marxist freaks in the EFF. I'm also not going to believe that the FSF are the only group on the planet who care about the prevention of such...and it nauseates me that you try and feed people such utter shit. If nothing else, at least try and remember that BSD UNIX (yes, in open source form, no less) existed for more than ten years before the GNU project. For fuck's sake, go and read about the initial history of UNIX while it was first being developed, if you haven't already. Stallman had NOTHING to do with that while it was inside AT&T, and those developers had the idea of freely sharing source from the word go.
The FSF do NOT have a monopoly on the desire for freedom...that's another thing Stallman insists on taking credit for that doesn't belong to him...I've noticed he's really good at that. It's just one more thing that he and Microsoft themselves have in common.
As long as Tivo or your hypothetical router manufacturer releases the source code to any changes they make to the GPL software they use, there's no problem. You're continuing to receive the benefits of whatever improvements they make to your code. That's what the GPL's supposed to be for. Businesses trying to make money off of GPL'd software will try all kinds of business models. Some will succeed, and some will fail.
Nobody's forcing you to buy their hardware, and if they want to sell loss-leader hardware using GPL code, that should be their business. Just because you *want* to take a cheap router and use it as a general purpose computer (or whatever) doesn't mean you have that right. And just because that piece of hardware uses GPL software, that doesn't change everything.
Now, if Dell started selling computers that ran 'Dell Linux' and couldn't be upgraded, I guess that would be a bad thing. But such a computer ought not to succeed in the marketplace. What's more likely is that they'll start selling ultra-DMA-infested Windows computers that can't run anything buy Windows. But the GPL won't help you there, either.
Posted from my Android phone. Oh, I can change this? There, that's better...
Microsoft wants sweet gpl'd code, yet they don't want to give source back, can they do this?
No, the gpl prevents them to.
Ok they will give source back but will patent it to prevent other developers from using it, can they do this?
No, they have to give them the rights to their used patents.
Ok what if they pay a third party company to develop it instead?
No, if 3rd party is granted a patent license they have to extend it to their users.
Ok, what if Microsoft does NOT grant the license?
Not practical, 3rd party will be acting illegaly and be exposed to prosecution, even worse Miscrosoft could be acused of not enforcing its patents, worse case scenario, they could even lose the patent.
Ok, what if insted of granting a license Microsoft makes a non binding reversable PROMISE not to sue and dress it up as a comercial, reciprocal deal?
Yes, this will allow 3rd party to use MS patents in their extended version of gpl'd made by the comunity, giving away code but not patent licenses since they didn't have the license to begin with yet be legaly protected by the deal, while Microsoft can't be accussed of selectively enforcing their patents since they dressed it up as a cross-patent deal.
Net result, Microsoft get's to distribute gpl'd-yet-non-free code, 3rd Party get's to be the last player in the Linux arena.
Who want's to play 3rd party?
It's also an object lesson on the pitfalls of lying one's ass off about everything. Sure, IBM had the lawyers, but SCO's case never had merit to begin with.
Which is why SCO can't back out gracefully.
IBM's policy (AFAIK) isn't to necessarily grind the opponent into the dust. It's more like "billions for defense, not one cent for tribute." They can accept the opponent throwing in the towel and retreating with some of their company intact.
(Example: Xerox's foray into the mainframe IBM-clone business: After Xerox decided it was failing, folded it, and sued IBM for unfair competition they proved in court that Xerox's business had actually been wildly profitable - but their accountants hadn't understood how to handle a leasing business so it LOOKED like they were losing money. Xerox went away with egg on their face and no mainframe business, but they're a power in the business machine market even today.)
But faced with an opponent who can't back down, the only effective way to carry out the policy is to burn their campus and sew salt on the ruins.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
> ... why bother... their stock is toast, so couldn't IBM just buy a controlling
> interest for $11.2M and wind it down?
No. They tought of such a turn two years ago and built into the statutes voting restrictions for new shareholders with more than 15% of stock.
Just from reading the GPL, I can tell that the NVIDIA driver is obviously not a derivative work of the Linux kernel. A derivative work requires that you take the source code of the kernel and modify it, which the NVIDIA driver obviously doesn't.
:-D
You might argue that a module is an extension of the kernel. Merely extending the kernel is no modification if a plug-in interface exists (so that the original source is not modified). No matter if done at source level or loadable module level. If you don't modify a single byte of the source code, it's not a modification.
A kernel module is an application program that is loaded by the kernel (a kernel-level application), and hence, the kernel isn't touched in any way whatsoever, by the kernel module. If that's the intention, to make code that runs in the kernel subject to the GPL, then the GPL is utterly ill-defined for that purpose.
I've been writing free software since the 1980ies, even before the GNU project existed, and hence I've been dealing with licenses for many years now. I created my own licenses, which were derivative of popular licenses that existed back then, which were similar to the BSD license.
If that NVIDIA case ever goes to court, you will see for yourselves that the claim that a kernel module comprises a derivative work of the kernel is utterly unfounded.
(BTW, I bet that NVIDIA's lawyers have already sorted the whole thing out before they published their first Linux driver.)
It's funny seeing GNU/Linux people spreading FUD, I always thought that was a Microsoft priviledge.
David Gould
main(i){putchar(340056100>>(i-1)*5&31|!!(i<6)<< 6)&&main(++i);}
Hi, my name's Ron. I've come under a lot of criticism lately, and I think it's unfair. My friend Steve and I have come to an agreement that we both think will help our customers, but some people seem unhappy about it. It's really rather a simple deal. I promise not to beat up anyone who gives Steve some money, and Steve promises not to beat up anyone who gives me some money. I give Steve some money, and Steve gives me some money. Everybody benefits from this; both the people who don't get beat up, as well as Steve and I. We've had a close look at our agreement, and we can't see that we're doing anything wrong. It's not like I'm threatening to beat someone up if they don't give me any money. No, it's Steve that's threatening them. I'm just saying that if you give me some money that makes you my customer, and Steve has promised to not beat up any of my customers. My lawyers have looked at the agreement, and they've said there's nothing illegal about not beating someone up. I know that some people became upset when they heard about our deal, but I think they're just over reacting. You've got to realise that it's just business, and that means that it's really OK. People do things like this in business all the time, so it's not fair to complain when I do it too. I know that in the end, you'll all get over this. I'm looking forward to you being my customer and in a way, Steve is too.
How we know is more important than what we know.
There's a really simple test. Does nv-kernel.o run without a kernel? No? Then it's a derived work of the kernel. That's the test that courts use. You really think this is the first time anyone has thought about copyright on plugins or modules? According to what you're claiming, Microsoft could write extensions to Java and distribute them seperately with a script to link them in at runtime and Sun couldn't do anything about it. Guess what, they had that court case last year, Sun won.
How we know is more important than what we know.
Sure there are. Check "man 9 intro" and the output of "ls /usr/share/man/man9".
Or look at Linux Kernel API....
They are #including header files from the kernel. It's clearly a derived work.
Does a userland C program which does a "#include " become a derived work of the standard C library?
If the same exact source code works on both Linux, Windows, and BSD, does that C program somehow have to be a derivative work of Linux's GNU libc, the Windows/Visual-C libraries, *and* the traditional BSD libc, all at the same time...?
"The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
That's supposed to be "#include "...I forgot to escape the less-than & greater-than symbols in HTML posting mode.
"The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
Does a userland C program which does a "#include " become a derived work of the standard C library?
YES!
That's why the standard C library is under the LGPL, not the GPL.
Jesus, it's like copyright 101 in here.
How we know is more important than what we know.
People like being confused, I guess. :-)
You don't have to be a lawyer to understand that GNU libc is released under the LGPL, and if it were released under the GPL instead then any program using printf in glibc would be required to be GPL as well.
Kernighan and Richie would certainly disagree with that claim.
I'll repeat an observation I made to someone else: if the same exact source code works on both Linux, Windows, and BSD, does that C program somehow have to be a derivative work of Linux's GNU libc, the Windows/Visual-C libraries, *and* the traditional BSD libc, all at the same time...?
"The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
You're comparing apples to pears. Java wasn't GPL'd back then. Sun has their own license, which I haven't read yet. Where can I find details about the case?
And what would prevent anyone from shipping extension libraries to Java? Companies are doing that all the freaking motherforking time when distributing their Java applications (often including the full JRE or even JDK).
I guess, unless there's a precedent, we don't know for sure how such a case would be handled. Namely, in the real world it doesn't just depend on the text, but also on its interpretation by lawyers and judges.
Seeing how ambiguous the GPL is, I might not use the GPL again for further works of mine (or wait for a GPL version which is crystal clear), and I won't recommend using it to anyone.
Also, I wonder what the use of GNU/Linux is when companies can't even distribute closed-source code for it without getting criticized. This doesn't help GNU/Linux become an industry standard; in fact, it will return it to an ever smaller getting niche when corporate benefits of it turn out be zero.
Let's say, company XYZ publishes a kick-ass (but properietary) database system for GNU/Linux named FGH. Everyone from the GNU/Linux world goes "Waaa! FGH isn't open-source! It's not LEGAL! It's DOOMED!" then company XYZ will get the impression that marketing or developing the product FGH for GNU/Linux was a mistake, and promptly withdraw it.
This would go on for all eternity until no more software is written for GNU/Linux.
(It is already not very programmer friendly, I must say!)
K&R can not disagree with license terms of a library that you agree to - regardless of the function name. If you do not agree to the license terms of anything, do not use it!
The answer to your question is "No" it does not. Why? because the licenses for all those libc libraries specify that they do not - including glibc which is LGPL'd not GPL'd.
If a library was GPL'd and you link to it, then your program is by definition a derivative work. If you do not like this, don't link your code to GPL'd libraries or the GPL'd linux kernel.
--jeffk++
ipv6 is my vpn
Actually, no, the standard C library was defined by Kernighan and Richie. It was originally implemented by them back in 1971 as part of AT&T V1 Unix, which was later made available in the mid-70's in what is now called BSD Unix. GNU Libc is a GPL'ed re-implementation of Kernigahn & Richie's published APIs, created by Roland McGrath around 1992.
Jesus, it's like copyright 101 in here.
Yep, and I've got news for you: you're failing to pass.
If you think that relicensing the GNU libc under the GPL rather than the LGPL would make every C program which does a #include <stdio> a derived work of GNU libc, then you don't understand either the history of the C language or copyright law.
"The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
Sigh. This is why the C library is under the LGPL, not the GPL, so people can make proprietary applications for Linux. It's a strategy thing. As for making extensions to Java in the form of jars or class files, the license for Java specifically allows that. Much like the license for the C library specifically allows that. Much like the license for the kernel specifically allows the creation of proprietary software that uses the syscall api to access the kernel. Copyright is really simple ok? You have no rights unless the copyright holder has granted them to you. Sure, there's fair use and such, but that's the exception, not the rule, and no, you're not going to use fair use to get away from being a derivative work.
How we know is more important than what we know.
I've never had to agree to the terms of a library simply in order to use it.
I might have to agree to those terms to redistribute the library itself, but, per the GPL clause #2:
"These requirements apply to the modified work as a whole. 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."
If the same source code can be compiled using a BSD libc and a GNU libc, then that source code "can be reasonably considered independent" and the GPL does not apply to the program itself.
"The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
I assumed you were talking about glibc, sorry about that.
1. If you link your code to glibc, you are making a derived work of glibc.
2. If you distribute your source under a restrictive license without linking it to glibc and you only use apis that were documented elsewhere, then you're not distributing a derived work of glibc.
3. If someone then links your source with glibc, they are creating a derived work of glibc, which they may then distribute because glibc is under the LGPL, not the GPL.
4. If, however, you were to use glibc specific apis or extensions to the elsewhere documented apis, then you are making a derived work of glibc. This is the so called "independant programs" test. Does your program require glibc to run or will it run with something else? Requires glibc? It's a derived work of glibc.
So let's talk about GNU readline instead of glibc. It's under the GPL, not the LGPL.
1. If you link your code to GNU readline, you are making a derived work of GNU readline.
2. If you distribute your source under a restrictive license without linking it to readline and you only use apis that were documented elsewhere, then you're not distributing a derived work of readline. Problem is, there's no "elsewhere" for GNU readline. It was invented there.
3. This is moot as you can't legally distribute the source, let alone a binary.
4. And yeah, moot again because your program clearly isn't independant. You could remove the readline specific parts and distribute that under a restrictive license, but that's irrelevant.
So getting back to what we were talking about.. NVIDIA have written a kernel module which contains some source portions and some binary portions. They provide the source portions so the community can maintain their code for them. The whole thing is under a restrictive license. The situation is identical to the readline case. They're using apis that are a part of linux and are documented nowhere but in linux. They didn't exist before linux. They are not co-incidentally compatible with linux. NVIDIA has deliberately set out to create a restricted work that links to GPL code. A blind chipmuck with a pencil in its teeth could win this court case.
How we know is more important than what we know.
But if that code links to the Linux kernel and requires it to function, the combined product isn't all your own code anymore, and therefore requires that you have the correct legal permission to link it. Abiding by the GPL is the only way to get this permission, therefore, if you want to link to the Linux kernel your code must be GPL-compatible too!
"[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz
You deserve credit for acknowledging at least the most basic point. :-)
1. If you link your code to glibc, you are making a derived work of glibc.
This claim is still very much open to question and is, as yet, an untested matter of law, when the library in question is a standard part of the operating system.
2. If you distribute your source under a restrictive license without linking it to glibc and you only use apis that were documented elsewhere, then you're not distributing a derived work of glibc.
Yes, agreed. This is one of my main points.
In fact, you could dynamicly link to a shared version of libc.so, and distribute your binary without including libc.so, and your program would still not be a derivative work, so long as it was using publicly documented APIs found elsewhere.
Let's skip ahead to the GNU readline case:
2. If you distribute your source under a restrictive license without linking it to readline and you only use apis that were documented elsewhere, then you're not distributing a derived work of readline. Problem is, there's no "elsewhere" for GNU readline. It was invented there.
There are actually at least two libraries similar to GNU readline which are not under the GPL; one was written by Christos Zoulas and is called editline (or libedit.a) and ships with FreeBSD, NetBSD, and OpenBSD, I believe; and there was another implementation called "BSD libreadline" implemented by Jaromir Dolecek which appears in NetBSD and MacOS X 10.4 (aka "Tiger").
3. This is moot as you can't legally distribute the source, let alone a binary.
4. And yeah, moot again because your program clearly isn't independant. You could remove the readline specific parts and distribute that under a restrictive license, but that's irrelevant.
There are proprietary programs which will dynamically load libreadline, if it exists, or use their own internal readline/line-editting capabilities otherwise. So long as the program works in either case, and works with the BSD licensed implementation of readline, the program "can be reasonably considered independent" of the GNU version of readline. Per the GPL clause 2, that program is not considered a derivative work of the GPL'ed library and "this License, and its terms, do not apply to those sections when you distribute them as separate works."
So getting back to what we were talking about.. NVIDIA have written a kernel module which contains some source portions and some binary portions. ... ] The whole thing is under a restrictive license.
[
Agreed.
They're using apis that are a part of linux and are documented nowhere but in linux.
You sure about that?
A review of the driver sources and the use of nm or otool to look at the symbols of the binary portion suggests that the driver is using primarily nVidia's internal APIs and then those for MESA/OpenGL's GLX extension to X11, not Linux specific APIs. Go look for yourself.
They didn't exist before linux. [ ... ] NVIDIA has deliberately set out to create a restricted work that links to GPL code. A blind chipmuck with a pencil in its teeth could win this court case.
Actually, X11 and I believe the GLX extension predates the existence of Linux.
nVidia itself doesn't quite, but nVidia was releasing proprietary drivers for Windows for nearly a decade before they released a Linux driver. The binary portion of the nVidia Linux driver is not a derived work of Linux or the Linux kernel, and you can check for yourself that the binary portion uses no symbols specific to the Linux kernel-- only the shim they've published in source code form does. The end-user of a system is free to compile that shim to produce a Linux kernel module which can be loaded, but, because the GPL and the nVidia proprietary license conflict, the end-user cannot redistribute the combination, which is a derivative work of both.
"The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
That is a valid point, however in order to reduce the confusion, glibc is distributed under the LGPL:
also, just to be certain: having a commonly named function like 'printf' in my library does not mean that you are free to use it in a proprietary product unless you comply with my licensing terms!
--jeffk++
ipv6 is my vpn
"Much like the license for the kernel specifically allows the creation of proprietary software that uses the syscall api to access the kernel."
Well then, where is this miraculous license that you're talking about? I only see GPL V2, and it doesn't mention the syscall API.
And if that's true what you're saying how is that compatible with the statement
"you're not going to use fair use to get away from being a derivative work."
So, what is it that makes the NVIDIA kernel module a derivative work of the kernel? Be more specific, otherwise it's hard to follow your train of thought.
A review of the driver sources and the use of nm or otool to look at the symbols of the binary portion suggests that the driver is using primarily nVidia's internal APIs and then those for MESA/OpenGL's GLX extension to X11, not Linux specific APIs. Go look for yourself. No, we're talking about the kernel driver here, not the X11 module. This is the stuff that gets installed in
How we know is more important than what we know.
I have trouble believing that you can follow your own train of thought. I'm really through talking to you. I said this to you before, this is a complicated conversation we're having and there's certain things that we expect participants to at least have a basic grasp of before they try to join. Probably number one on that list would be the syscall API exception to the GPL on the Linux kernel. If you don't even understand that then WTF are you doing here? Go read up on the subject before you open your mouth.
How we know is more important than what we know.
Ah, you're a troll. I didn't get that right away. I thought you were serious there for a moment.
fuck, you're an idiot.
I hope you are sterile.
How we know is more important than what we know.
Well, first you should learn what a linker does and what a C compiler does. To baby you a bit there: - A C compiler is a program that translates a source code module written in the C programming language (.c) into a machine-readable form, most often a machine code object module that has external references to libraries (.o). - A linker combines these object modules into an executable program with or without externel references to shared libraries (.so). A shared library might also be linked to at runtime. Hence, source code has no linkage. When a program is compiled, it becomes linked to whatever system environment that is used by the development system (i.e. compiler, linker, libraries). Hence, a program in source code form cannot be a derived work of a library that it just envisions to use. In the source code form you do not necessarily know which environment there is for your program after compiling. A binary program having external references that link at load time therefore is not a derived work of its environment, because the environment is resolved at load time, and not at compile time. Furthermore, you can compile a program on one platform and move the binary to another.
The source code of the driver sits there during compilation of the kernel module, it does not modify the kernel.
I don't understand the difference between GPL v2 and v3 at all. But I'm beginning to think the difference is irrelevant. From IBM's rebuttal submission to SCO:
-------------
8. The intellectual property rights of GPL licensees or others may not be used
to "overwrite" or create an exception to the restrictions of the GPL.
(Ex. 128 7; Ex. 129 & 11.) Section 7 of the GPL expressly states:
If, as a consequence of a court judgment or allegation of patent
infringement or for any other reason (not limited to patent issues), conditions
are imposed on you (whether by court order, agreement or otherwise) that
contradict the conditions of this License, they do not excuse you from the
conditions of this License. If you cannot distribute so as to satisfy
simultaneously your obligations under this License and my other pertinent
obligations, then as a consequence you may not distribute the Program at all.
(Ex. 128 7 (emphasis added); Ex. 129 11.)
--------------
From the LGPL:
--------------
If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Library at all. For example, if a patent license would not permit royalty-free redistribution of the Library by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Library.
--------------
IANAL but to me it seems that the GPL and its variants already protect against patented software. If Novell release something under the GPL then we can assume that it doesn't infringe MS's patents, because they are in the best position to know for sure having signed this agreement and presumably seen the evidence. If they want to release OO.o with MS patented stuff in then let them - by releasing it under the LGPL they have either broken the license, in which case, sue them if Microsoft sue you - or they haven't, in which case Microsoft haven't a leg to stand on. And if they don't release it under the LGPL, then anyone who has contributed to OO.o can sue Novell for copyright infringement. No?
I get more and more the impression that people like you overestimate the scope of the GPL. An isolated code module that the compiling environment decides on -- in most cases, the configure script which decides which libraries are used for source code inclusion and later linkage by the development linker or system linker -- does not, ever, imply a derivative of the surrounding environment. That's plain nonsense.
I told you to fuck off asshole. You can't bully people into talking to you. I tried to be nice. I tried to explain things to you, but you just don't know what you're talking about, now go away.
How we know is more important than what we know.
Yes, what you describe is plain nonsense, and that is not what I am saying. You are confusing the tools with the libraries. The mere act of compiling does not have any affect with the GPL - Where it matters is when you distribute the result!
Are you saying that if my configure script for my closed-source application accidentally selects the 'free' version of the trolltech QT library, then I don't have to pay the QT license fees? If so, then it appears that I can save $6,000.00! woo! Somehow, I feel it does not work like that.
It all comes down to this: If I write a library and have a license that says you must give me $1000.00 for every time you ship a product using it, you can either pay me or you can not ship product that is linked to my library - even if it is a libc or libm replacement! Especially considering that my personal libc library runs on obscure embedded DSP chips and my libm is fully vectored!
Similarly, if I write a library and have a license that says that if you link to it then your program must be GPL'd, then you must GPL your program if you distribute a binary of your program linked to my library - even if all you use is my special GPL'd version of printf().
It is your responsibility to make sure that the closed source binaries that you distribute are legally unencumbered - it is not autoconf's job.
--jeffk++
p.s. and for others reading, glibc is under the LGPL, not the GPL.
ipv6 is my vpn
What I want to know is, where in the GPL those "viral" statements are that have been mentioned. What makes the NVIDIA installer GPL-encumbered? That's all I want to know, and no "it just is" please.
Why, thank you!
I scrapped Linux on my machine and installed Solaris. No more GNU in the kernel, it's that simple!
When you refer to "IP" in the context of software patents you should always use quotes to make the point that all the system is bogus.
Repeat after me: software patents are nonsense. Anybody claiming to have one is just milking th broken US patent system.
IANAL but write like a drunk one.
.... to deal with companies with a suspect reputation?
We see it here all the time, people having absolutely no moral qualms to deal with MS and defending them like if they were a role model of ethical behaviour.
And now, the above. If everything is working, who cares!
Well, I do. I do not want to deal with companies that do not meet my standards of integrity.
Although I am still in the fence regarding Novell (and thus SuSe) frankly they have not endeared themselves much.
Showing the willingness to do such a deal, with a company with the patently bad reputation as MS, should tell us a lot about Novell.
I don't deal with convicted entities, for bunnies sakes, they could screw me, I know they have done it with others and convicted for that. Why should any sane person or company deal with them? And why should I keep a bussiness relationship with somebody that is dealing with companies with a shoddy reputation?
No, no, and no.
IANAL but write like a drunk one.
Hi Nick,
I agree with pretty much all you've said and am glad that you have made the distinction, that maybe I've been unable to express properly, between SuSE and Novell.
This is basically what I'm trying to get at; the arrangement between MS and Novell is a corporate contract and, from personal experience of over 35 years in business (I've been a London based director of a unit of a US multinational, retired and then formed my own business), it is definitely the norm to have NDA's within non-compete/joint-dev projects.
The problem for the the (F)OSS community is that they are used to openness as a matter of course and suspect anything that appears to be hidden, the fact that MS are involved magnifies that suspicion into stratospheric realms.
Whatever parts of the arrangement between Novell and MS are thus covered will not be released to the community or the industry unless one or other of them decides to break the contract - the (F)OSS community will have to live with that.
By the nature of GPL, anything that is released to the SuSE (not Novell SuSE Enterprise clients) community has to be under a GPL license and, personally, I cannot imagine SuSE maintainers accepting anything that even smells faintly tainted unless it is headed up with a GPL license co-signed by Novell and MS.
As I've said elsewhere, the spotlight is on this deal, the players and their actions and it's not going to be turned off in a hurry. Any code being released from the joint develpment will be scrutinized to the nth degree - by just about the whole (F)OSS community. To say beforehand that "the code will be tainted", "there'll be MS proprietary code injection" etc., etc., is rather jumping the gun don't you think?
Wouldn't a better approach be to say, "Ok, we don't trust this covenant, however that's a business decision that you (Novell and Microsoft) have agreed to, but should you release code to the (F)OSS areas then it must be released under GPL and co-signed as such by both of you."?
Have FSF check or write the license if that's what makes people comfortable but don't just reject everything out of hand.
I think that anything that aids interoperability between (F)OSS and proprietary softwares (particularly MS software) is a good thing for Linux - neither are going to go away in a hurry - and is also a good thing for MS.
Your analysis of how MS looks at the future is, I would say, correct but with an additional element, MS has seen the direction that the world is going and it threatens their longevity if they do not actually adapt to some of those changes.
The belated entry of MS into the web-sphere (admittedly, they tried before with the limited Passport and Spaces/Groups concept - they din't grasp the idea of it being an uncontrolled area that does not fall into a corporate model) has shown both MS and the rest of the interested tech' world that they, a) are not currently very good at it (but they are learning fast) and, b) they have some serious catching up to do (for which they are going to need strategic partners).
MS are learning about "Adapt or Die", they as sure as hell don't like it (and there's a possibility that the likes of Steve Ballmer may never get it) but it's a fact of their life. The other thing that must have really stung MS is that this area, like FOSS (FSF), cannot be bought. The one area where ideas are not for sale just happens to be the future of their industry.
MS will have to work with (F)OSS, previous competitors and new competitors to survive and their "embrace, extend and extinguish" methodology can no longer work - it isn't possible to "EE&E" a social idea or movement.
There are other points that I would bring up but they've been written elsewhere on /. so I'll pass on all but one:
If SCO taught the market anything at all, it is that suing your own customer base is not a viable business model. For MS to try and sue anybody other than corporate competitors would be an act of suicide, would o
Democracy is being able to elect your own megalomaniac, a dictatorship cuts out the middle man.
If MS wanted interoperability they would not need to sign patent bullshit.
They could just sit down, say here is how our stuff works, lets make it work with Linux.
Even if they wanted to sort out the patent bullshit and were acting in good faith, they could talk to the different Linux providers to address the issue *before* getting ready for litigation.
As for MS stuff making work their wares with Linux, gives us a blessed fucking brake, in the immortal words of Johnny Mac: "you cannot be serious!".
MS is going to loss lots of face and top dollar (at a time whne their shares have been distinctively unimpressive of late) if the mess around much with this.
IANAL but write like a drunk one.
Thanks for a challenging and thought provoking reply. As you say, I think we have a lot of common ground here. All the same, I'd like to explore a few issues a bit further.
I'll grant you the lack of openness in this agreement is a major cause of friction; I'm just not sure it's fair to characterise it as the FOSS community's problem. Novell have a lot to lose if they lose the support of the community.
And I'm not sure we do. Certainly we'd be the poorer without Novell's contribution, but as you say, it's hard to kill off a social movement. I think Novell needs us more than we need Novell, and from that perspective I think Novell have handled this deal spectacularly badly. They've done a deal with a notorious untrustworthy software corporation, and then left themselves no way to demonstrate good faith except to say "trust me".
I think that probably qualifies as an error of judgement on their part. FOSS is used to being lied to by corporations, especially when MS is concerned, and "trust me" just isn't going to do so far as a lot of people are concerned.
I think the problem is that Ron Hovsepian and the top level management at Novell still don't understand the community. So now people want to know how wide the gulf is, and anything Novell say in their defence is going to look like MS style spin.
mmm... but this isn't about proprietary code, it's about patents. Look at it this way: fundamentally, the whole thing is smoke and mirrors. Microsoft has a whole arsenal of loosely worded patents it could probably bring to bear any time it wanted. It would cause a dreadful nuisance, and take a lot of beating back, as you say, a likely end results would be the invalidation of a most or all of those bogus patents.
On the other hand, say there's something clear and narrowly scoped, that, under current law would give them a clear and indisputable win. If they could get those patents included in mainstream Linux, they could cause the same amount of trouble for comparatively little risk. Thus for example the concern over the open office plugin; MS have been talking about having the XML schemas patented for a long time. But if Ron Hovsepian says "Open XML Compatibility" then there's a fair chance that MS patent violations will end up in SUSE, and from there in a lot of Linux distros.
It gets worse with Mono. If the word comes down on top to assure compatibility or interoperability with some feature that is actually covered by Submarine Patent X, then the whole of Gnome could be knocked back two to three years or more. Wouldn't a better approach be to say, "Ok, we don't trust this covenant, however that's a business decision that you (Novell and Microsoft) have agreed to, but should you release code to the (F)OSS areas then it must be released under GPL and co-signed as such by both of you."? If MS will also agree to open all relevant patents, unambiguously and without being limited to SUSE customers, I don't think anyone wil have a problem. I can't see that happening though. It's patents, not code, that is at issue here.
Not at any price. I'm all for interop, but we don't know what the hidden costs of this agreement are, and as you point out we'll likely never know. That makes it kind of a pig in a poke, and I find myself oddly loathe to invest.
Not yet
Don't let THEM immanentize the Eschaton!
Hi Nick,
I appreciate your dialogue and will try to answer your points, though more as a collective item than one by one.
Novell does indeed have a lot to lose from withdrawal of support from the (F)OSS community, equally the community have a lot to lose by ostracizing Novell or, more importantly, the Novell SuSE maintainers. Again, I'm trying to deliniate the two entities as I see them, Novell (the communications software company) and Novell/Open Suse.
The agreement with Microsoft is between the former and, by employment, some of the latter. The corporate entity can and will make commercial decisions that do not necessarily take into account the wishes or expectations of the employed maintainers let alone non employed contributors. The board of directors have a duty to shareholders before any other interests, the (F)OSS community do not have that constraint but have happily worked with the Novell and open SuSE maintainers for some years. That relationship should have built some trust for the SuSE guys, employees of Novell or otherwise, and that should be the bellweather test for the MS/Novell agreement.
The moment that the maintainers feel unhappy about the deal and start to vocalise or leave because of it should be the point at which the rest of the linux community shut the door to Novell.
Look at it this way: the code within Novell SuSE is examined by both the in-house and volunteer groups. The volunteer group don't allow any proprietary code to be directly incorporated into the OpenSuSE branch - the repositories are there for people to use if they want but the core offering is vanilla. To be fair, the Novell-SuSE distro uses the external repositories too - compare this to some of the other distros that have proprietary graphics drivers included within their "vanilla".
For the sake of argument, I use proprietary to mean patented or copyrighted code - the difference is not too relevant as both can result in a law suit for breach or infringement - the supposed "methods" that have been patented are, as I think most sane and savvy people would agree, mainly bogus.
As I said before, this argument is largly a domestic US one, software patents may stand up in a US court but almost certainly would not do so in the EC (Microsoft's largest market) and an attack upon (F)OSS would be seen as an attempt to manipulate the market by destroying competition. I'm not sure that Microsoft would want that case to be brought against them having promised to reduce their lock-out ethos within the EC area (and having been fined for not complying fully or on a timely basis).
Regardless of what the McCreerys et al may wish, software patents in the EC are not allowable at this time - the reason no-one has dared to try and go to court and prosecute an "infringement" is precisely for that reason, the patents would be ruled invalid under current EC law whether granted by the EPO or any national PO or not, they are excluded from patenting.
Novell, as distinct from SuSE, has made a business agreement with Microsoft. The minutiae of that deal will probably never be opened to the public domain unless there is an escape clause in it somewhere, which I doubt. However, it's almost certain that, if it was Oracle, IBM or anyone else who had made a deal the same would pertain. Just because Novell is the nominal owner of SuSE Linux does not mean that corporate Novell is going to be an open book to all and sundry, it is a for profit company and will continue to operate as such and for the (F)OSS or anybody else to expect any different is unrealistic.
Novell, however, are not the guardians of SuSE Linux, the employed and volunteer maintainers have that distinguishment. Those are the people with whom my trust rests and a lot of the community seem to have forgotten that fact.
My "cry wolf" comment is very much intended to reflect the way the community is percieved outside, by people who are not developers (or /.ers), the potential customers, allies, press and plain uninformed. For those people
Democracy is being able to elect your own megalomaniac, a dictatorship cuts out the middle man.
You seem to be suggesting that because business is getting increasingly involved with FOSS, it's the responsibility of the community to adapt to commercial procedures. That seems to be something of a meme in the community: The idea that somehow we are all desperately in need of corporate involvement, and that unless we bend over backwards to keep industry happy, our efforts as amateurs and enthusiasts are sure doomed to failure.
I'm not ready to accept that. It seems to me that FOSS made a useful operating system from scratch with little or no corporate backing, and frequently in the face of corporate scorn and opposition. and that only once we had something useful did business begin to take an interest.
We have something that many business want. But if they want it, they will have to deal with the community on our terms, not theirs.
I can accept that Novell is following standard business practice, and that's all very well if the deals pertain to NetWare, or something else for which they own all the rights. On the other hand, they are making agreements concerning software which they do not own, and which concerns lawsuits aimed at the other distributors, developers and users of this software. Furthermore this deal involves getting a considerable sum of money ostensibly in exchange for something Microsoft neither wants nor needs. Somehow it doesn't seem at all reasonable to say that this is none of the community's business.
Free Software is a relatively new thing. Some established business methods aren't going to work where Free Software is involved, and I think we need to evolve new ways of doing business in this sphere. Like you said, it's a case of "Adapt Or Die", and that goes for Novell-The-Communications-Software-Company just as much as it does for Microsoft.
I value corporate involvement, but again, not "at any price". And if Novell-The-Communications-Software-Company can't get its corporate head around that idea, well then I'd sooner leave the dessicated remains of the company hanging from a gibbet than I would have us all sit around the campfire singing kum-by-ya while Ron Hovsepian and Bill Gates sell us all down the river.
mmm... but this is where the patent distinction becomes important. If the attack involves patents, then its entirely possible that there could be nothing to detect. Suppose for a second the upper echelons of Novell are up to no good. All it takes then is for Mr. Hovsepian to say "make SUSE compatible with Feature X". SUSE devs on the Novell payroll do as instructed, the SUSE scrutineers see no problem, and the feature is passed. Then, six months or a year down the line, MS announce that Feature X uses protocols or schemata or whatever that are quite clearly patented under US law, and starts suing the non-commercial distros. That, by the way, is why I take little comfort from the fact that EC law prohibits software patents; I think MS could do tremendous damage to the community purely by suing Stateside distros and developers.
In any event, I think it's unrealistic to expect the FOSS community ever to act with a single mind. Some people, some groups may well ostracise Novell based on this episode; others are likely to maintain good relations with them in the face of anything short of a direct attack. Most are going to be somewhere in between. The question is to what extent Novell has damaged the trust they worked so hard to build between themselves and the community, and whether they have left any avenues open whereby they might undo the damage. I think it's going to take more than a web page full of spin to do the job.
Don't let THEM immanentize the Eschaton!
Hi Nick, :) but I must take issue with your comment: You seem to be suggesting that because business is getting increasingly involved with FOSS, it's the responsibility of the community to adapt to commercial procedures. That is not what I was (or intended) saying, in fact the reverse is my viewpoint. FOSS (Ok, I'll omit the brackets) is becoming more and more involved with the corporate sector and there is a need for the community to be able to engage that sector in a language that they understand. The GPL was a "corporate" move by itself, it may have been a protective move but it is a move that businesses (particularly in the US) understands.
I'm glad that my "heretic" level has been reduced a touch
There is no disputing that the product is, literally, quite incredible having come originally from amateurs and enthusiasts, the fact that it is still maintained by like minded people is it's greatest strength. That does not diminish the need for the FOSS and commercial world to be able to communicate in a way and at a level that both understand.
I would suggest that the FSF could provide that service, the legal side is there but, to date, the PR side is not.
There is a need for a unified voice of FOSS, the message agreed upon by the community - no doubt with some compromises - and then issued as a statement on behalf of the FOSS community. The worst message that can be given is, "We're gonna get screwed and sued!", which seems to be the current stance, the message should be, "We think that someone is going to try and screw and sue us." and then give the reasons and suggested remedy. Putting the pressure on Novell and Microsoft and highlighting the potential problems to the world at large.
I don't see this as pandering to the corporate world. How many more people would be using a Linux system at home if at least one family member was using it at work?
If their employers are using it then it just adds to the relative "validity" of the Linux o/s and FOSS offerings so, to that extent, FOSS does need the corporate world.
I think that we will continue to digress over the injected code scenario and the relative merits of, a) anyone knowingly doing so and b) the real effects that it would have upon FOSS. Whilst I can quite understand your point about the damage done by US only patents litigation (or threat thereof), I would still suggest that the alianation that such action would engender, the regulatory interest and negative PR generated outside of the US plus the ability, by its nature, of FOSS development to relocate has the potential of being corporate suicide by the litigant.
In all other respects I agree completely with you and must say that I've enjoyed the discussion. In fact I don't think it unreasonable to say that this the sort of discussion that should be taking place between the FOSS community, FSF and corporate world now.
Jez
Democracy is being able to elect your own megalomaniac, a dictatorship cuts out the middle man.
I wouldn't have said "heretic". Aside from a slightly belligerent initial exchange, I've always found your position to be rational and sensible. We differ a little on how active the community should be when it comes to supplying Novell with feedback regarding their dealings with MS, that's all. In any case, I don't really consider myself a follower of St iGNUtious; I'm more a "linus" than I am a "richard" or an "eric"
Also, I must apologise if I mischaracterised your position. The intention wasn't so much to put words in your mouth as it was to clarify your position - but then I got carried away bashing one of my least favourite memes.
Certainly, any improvement in communications is always welcome, and I suppose realistically, if we sit around waiting for business to build a universal translator, we may be waiting for some time.
Not sure about the FSF in that role. They strike me as just a bit too political and a bit too anti-corporate to gain much credibility in business circles. I'd have though it was a better fit for the OSI. But then again, opensource.org doesn't have any news since July 2005 when they announced Michael Tiemann as interim president. So maybe not.
Of course, Bruce Perens has been doing this sort of thing off his own bat for a while now. Have you seen his petition, by the way? It's rational, polite, and fulfils most of your criteria for a community statement. It isn't as clear as some of Mr. Perens' writings, perhaps, but it gets the point across. And it has going on three thousand names to back the point, many of whom have some interesting points of their own to add.
Then there's the Samba team's open letter, which I think does a much better job of laying out the FOSS position and why we find Novell's actions so objectionable. Sadly, Novell have already read and dismissed this one. In fact, Novell's response across the board so far has been "you might think that, but this is why you're wrong". It makes a fellow wonder just how we can demonstrate how seriously we take this issue, except by voting with our feet.
I suppose... I think part of the problem is that Novell suffers from that disease which leads organisations to believe they can solve problems by altering the perceptions of others rather than by amending behaviour in themselves. Once a company starts to think like that, it becomes almost impossible to hold any meaningful dialogue with them, because any unwelcome opinions get routinely passed to PR as problems to solve, without any requirement that they pass through the brains of the people making the decisions.
It also tends not to work very well with the FOSS community. One of the first things you learn as a developer is to watch out for who can walk the walk, and who's just talking the talk.
Incidentally, speaking of the Samba Team, if you haven't seen it, check out Jeremy Allison's column in the latest Linux User and Developer. He goes to some pains to point out that as a Novell employee, he can't really criticise them in public, then points out that the Samba team is under no such prohibition, and provides a link to the Letter To Novell. He also finds some interesting parallels with the RedHat/Oracle dustup as well.
Agreed. And thank you - I've enjoyed the discussion too.Don't let THEM immanentize the Eschaton!