|
|
Log in / Subscribe / Register

Welte: Report from the VMware GPL court hearing

On his blog, Harald Welte has a report on a hearing in Germany regarding VMware's alleged GPL violations. Welte is a former kernel developer as well as the founder of gpl-violations.org, so he has quite an interest in the case, which was brought by Christoph Hellwig and is being funded by the Software Freedom Conservancy. To Welte's eye, it seems that there are two questions at issue: whether vmklinux and vmkernel are considered to be one or separate works (in a copyright sense) and whether Hellwig has the standing to sue: "This situation is used by the VMware defense in claiming that overall, they could only find very few functions that could be attributed to Christoph, and that this may altogether be only 1% of the Linux code they use in VMware ESXi. The court recognized this as difficult, as in German copyright law there is the concept of fading. If the original work by one author has been edited to an extent that it is barely recognizable, his original work has faded and so have his rights. The court did not state whether it believed that this has happened. To the contrary, the indicated that it may very well be that only very few lines of code can actually make a significant impact on the work as a whole. However, it is problematic for them to decide, as they don't understand source code and software development. So if (after further briefs from both sides and deliberation of the court) this is still an open question, it might very well be the case that the court would request a [technical] expert report to clarify this to the court."

to post comments

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 4:13 UTC (Fri) by xnox (subscriber, #63320) [Link] (3 responses)

Some features are thousands lines of code.
Some features are a single line.
Some features (bugs, and bug fixes) can be a single character (in a complex regexp), or even a single bit (feature flags when interfacing with hardware.

Every line of code used is significant, and if it weren't it wouldn't be used and/or disputed. I hope they get a decent technical expert. E.g. a programmer.

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 9:16 UTC (Fri) by justincormack (subscriber, #70439) [Link] (2 responses)

Yes but this is copyright law. One character may be highly significant but it is not copyrightable.

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 9:31 UTC (Fri) by wsa (guest, #52415) [Link] (1 responses)

Couldn't it be (at least in Germany) if this character is above the threshold of originality? I agree, though, for most characters, this is probably not the case.

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 10:13 UTC (Fri) by dakas (guest, #88146) [Link]

I've patched buggy binary drivers written in assembly language where the fix had to have a higher code density than the original. Very few bytes, high level of creativity.

Where should one, in contrast, place the click&drool web framework dingaloops? The stuff where it takes a better person than the clicker to understand the output (and likely throw 95% of it away for sanity reasons)?

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 8:18 UTC (Fri) by mm7323 (subscriber, #87386) [Link] (8 responses)

So if, from the article, the Linux parts are a tiny little yellow box, a company with the resources of VMware could surely reimplement those bits for themselves...

Aside from that, could the issue of whether Christoph owns enough 'unfaded' copyright not be easily resolved by a few more kernel devs jumping on board? Or do they not agree with the case?

The yellow box

Posted Feb 26, 2016 9:22 UTC (Fri) by wsa (guest, #52415) [Link] (1 responses)

From what I recall, they said they could replace the yellow box, but it would be a huge disadvantage in the market.

Personally, I have doubts about the first part, not so much about the second.

The yellow box

Posted Feb 26, 2016 11:43 UTC (Fri) by tialaramex (subscriber, #21167) [Link]

I too have doubts about them replacing the yellow box.

Probably the VMWare of 2000 could have done this. VMWare in 2016 is another huge dinosaur corporation. Such corporations have the resources but they don't have the focus of will to deliver such things. What you actually do if you run a big dinosaur like this is buy start-ups who have the pieces you need. But there isn't (and won't be) a start-up writing a Linux replacement, because start-ups go "Oh that part's solved" and use Linux.

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 12:18 UTC (Fri) by ukleinek (subscriber, #56625) [Link] (1 responses)

It would be great if the court could just say: "The kernel as is is licensed under the GPL and complex enough to be copyrightable" and then just decide if VMware is in conflict with the kernel's license independent of who actually holds the copyright (as long as it's not VMware which should be easy to prove).

Regarding other kernel developers that could jump on the board: I assume this would be helpful, but not every dev considers it important or shares the view that one should sue for it. Also officially "standing up", being involved in legal stuff, being obliged to attend a hearing can all be barriers.

Welte: Report from the VMware GPL court hearing

Posted Feb 27, 2016 2:32 UTC (Sat) by rahvin (guest, #16953) [Link]

Standing is a big deal in the courts, after all we don't want random strangers suing people in cases where they have no standing (ie no relationship that gives them a basis to sue) suing people. The courts would be unusable if such a thing were to occur.

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 14:44 UTC (Fri) by pbonzini (subscriber, #60935) [Link] (1 responses)

Christoph is among the top ten committers of drivers/scsi/ (and even higher if you exclude drivers). It may take some time to convince the judges, but it's just legal play whether Christoph owns enough copyright or not.

It's a completely different jurisdiction and even legal system, but in Google vs. Oracle the judge ruled that even *7* lines of code were not "de minimis".

Welte: Report from the VMware GPL court hearing

Posted Feb 27, 2016 6:47 UTC (Sat) by marcH (subscriber, #57642) [Link]

> It's a completely different jurisdiction and even legal system, but in Google vs. Oracle the judge ruled that even *7* lines of code were not "de minimis".

Does US / Common law know this "fading" copyright concept?

Welte: Report from the VMware GPL court hearing

Posted Feb 27, 2016 0:56 UTC (Sat) by vonbrand (guest, #4458) [Link] (1 responses)

The Linux part is almost all the guts of the product from what I understand, what is a tiny part (moreover much diluted by later work) it the plaintiff's contributions to the Linux pieces used.

Welte: Report from the VMware GPL court hearing

Posted Feb 27, 2016 2:31 UTC (Sat) by rahvin (guest, #16953) [Link]

I would not call the hypervisor a minimal part. Though it's of little value without all the drivers.

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 9:05 UTC (Fri) by wsa (guest, #52415) [Link]

I've been in the court room, too. Harald did a great summary IMO. Thanks for doing that!

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 10:30 UTC (Fri) by Sesse (subscriber, #53779) [Link] (14 responses)

This looks like it could become an interesting case. It's relieving to hear that at least, the court seems to understand and accept the concept of the GPL.

The part about standing to sue also is interesting to me; it might seem like an attempt to get the case dismissed on a (German-specific) technicality, but if it goes through, it will be practically very hard to sue for the specific case of the Linux kernel in the future.

/* Steinar */

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 20:28 UTC (Fri) by xtifr (guest, #143) [Link] (13 responses)

Standing is a pretty important issue in most courts. Not in the slightest a German-specific thing. You can't sue on behalf of some random other person in the US either.

And actually, if the case is dismissed for lack of standing, that still leaves VMware in a pretty vulnerable position, as someone who *has* standing for the code in question could easily come along and pick up the case at any time.

If I sued someone for violating Bob Dylan's copyrights, the case would be thrown out for lack of standing pretty much anywhere, but that would certainly not stop Bob Dylan from suing the same person on the same grounds.

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 20:31 UTC (Fri) by Sesse (subscriber, #53779) [Link] (12 responses)

The concept of standing in itself is pretty universal, sure, but the concept of fading of copyright appears to be German-specific. If Christoph doesn't have standing here, not a lot of people (or even corporations) would.

/* Steinar */

Welte: Report from the VMware GPL court hearing

Posted Feb 27, 2016 2:35 UTC (Sat) by rahvin (guest, #16953) [Link] (11 responses)

True, but they'll probably ask an expert to give them an opinion on it and you are going to have a hard time finding a single developer in linux probably even including Linus with a higher percentage of code that bears their name. I don't have the stats handy but I doubt a single person is responsible for more than 1-2% of the monster code base that is the Linux kernel.

Welte: Report from the VMware GPL court hearing

Posted Feb 27, 2016 10:39 UTC (Sat) by Sesse (subscriber, #53779) [Link] (8 responses)

Indeed, but the only conclusion from there doesn't have to be that he could have standing to sue. It could also mean that effectively _nobody_ has standing to sue.

/* Steinar */

Welte: Report from the VMware GPL court hearing

Posted Feb 27, 2016 11:33 UTC (Sat) by farnz (subscriber, #17727) [Link]

Third option; the expert could require joint plaintiffs to get anywhere, forcing Christoph to find another copyright holder who'd back him.

Welte: Report from the VMware GPL court hearing

Posted Feb 27, 2016 11:57 UTC (Sat) by Wol (subscriber, #4433) [Link] (3 responses)

Which would gut the licence.

Which, certainly in America, I understand Judges would be EXTREMELY reluctant to do. It's like taking an anthology of poetry, and saying that because ANY ONE author has only committed a little bit of work to the whole, that means that ALL of the authors have effectively given away their copyrights without intending to.

I can't see that flying, in pretty much any jurisdiction.

Cheers,
Wol

Welte: Report from the VMware GPL court hearing

Posted Feb 27, 2016 17:31 UTC (Sat) by tao (subscriber, #17563) [Link] (1 responses)

I think your comparison leaves a bit to be desired.

A single author can claim copyright on their work (their poem(s)), but not the entire work. That copyright is usually with the editor or publisher.

A more relevant comparison would be one of these writing projects where every person writes one sentence (or word, or paragraph, ...) and the next person continues.

Can one person then sue for copyright without asking the others? I'm not sure. It's a hairy issue, because no matter if you say yes or no you can get weird corner-cases.

Welte: Report from the VMware GPL court hearing

Posted Feb 29, 2016 11:11 UTC (Mon) by farnz (subscriber, #17727) [Link]

In this case, I would expect an English court (and probably an American court - their system is similar) to permit someone to sue for infringement over only the parts they supplied. In the event that the court says that there was infringement, an English court would then say two things:

  1. If any of the other copyright holders claim against the defendant for the infringement of their rights, the findings of this court mean that you are deemed to have infringed, and that you knew about it from the date of the judgement against you.
  2. You will only be compensated for your share of the work, not the full work - this may, however, include a requirement to stop distributing or using your share of the work.

Point 1 is the scary one from a violator's point of view - if VMWare lost this case in an English court, then every developer whose code they've used without authorisation could claim compensation. Point 2 carries other implications - for VMWare, they'd have to stop distributing ESXi and anything else with Christoph's code in while they rewrote it, and they could find themselves banned from using any version of the Linux kernel with Christoph's code in, making compatibility with Red Hat and SuSE a pain. This point is, however, why a minor copyright holder is a very different fish to Christoph - taking Linux minus my code, and paying a contractor to fill in the missing bits is trivial. Not so much if they have to remove everything Christoph holds copyright on.

That said, this is a German court, using a different legal system, so they may act differently.

Welte: Report from the VMware GPL court hearing

Posted Mar 6, 2016 3:49 UTC (Sun) by giraffedata (guest, #1954) [Link]

"It could also mean that effectively _nobody_ has standing to sue."

I think you missed the point of the fading concept. It doesn't mean Christoph has no standing because he wrote only a small part of the work; it means he doesn't have standing because nearly all of his work has been replaced. So if Christoph doesn't have standing because of fading, someone else certainly does (someone who modified the code after Christoph).

Welte: Report from the VMware GPL court hearing

Posted Feb 28, 2016 1:03 UTC (Sun) by rahvin (guest, #16953) [Link] (2 responses)

Which is so highly unlikely I don't even consider it a possibility. As other have pointed out, the sheer amount of his work in the Kernel could literally be several man years. Just because the kernel isn't so large doesn't mean no one has ability to enforce it's copyright in Germany. What I expect the court will find is that because of the sheer size of the kernel codebase that even 1% is a valuable and irreplaceable part of the kernel.

Welte: Report from the VMware GPL court hearing

Posted Feb 28, 2016 1:31 UTC (Sun) by Sesse (subscriber, #53779) [Link] (1 responses)

You may consider it not a possibility, but seemingly VMware is attempting it as a defense strategy, and the court “recognizes it as difficult”.

I am hoping it comes down on the enforceable side in the end, of course, but you can never be quite sure.

/* Steinar */

Welte: Report from the VMware GPL court hearing

Posted Mar 1, 2016 0:51 UTC (Tue) by rahvin (guest, #16953) [Link]

At least in American courts (not sure about German) any issue you don't bring up in court in the initial stages will be automatically excluded later. As a general rule the result is Lawyers will throw everything possible at the wall and hope one of them sticks.

Welte: Report from the VMware GPL court hearing

Posted Mar 3, 2016 10:19 UTC (Thu) by dunlapg (guest, #57764) [Link] (1 responses)

I'm not sure why "percentage" has as much to do with it as actual raw numbers. The Linux kernel has what, over 20 million lines of code? 1% of 20 million is 200,000 LoC. Even 500 LoC should easily be a copyright infringement; I'm sure Christoph has more than that.

I mean yeah, if you performed a 5-minute song, and then it was re-pasted over and re-edited until the part you actually performed was only 10 seconds long, then I'd call it fair use at that point. But if you performed a 2-hour concert, and then there was an edited version that only had 20 minutes by you, it doesn't make any sense to say those 20 minutes somehow are up for grabs because they were originally part of a 2-hour work.

Welte: Report from the VMware GPL court hearing

Posted Mar 3, 2016 11:11 UTC (Thu) by juliank (guest, #45896) [Link]

Fair use is irrelevant, as no such concept exists in German law.

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 11:20 UTC (Fri) by ledow (guest, #11753) [Link]

It seems to me that if the case really is coming down to two questions:

1) does Christoph hold sufficient rights on the code to bring forward the legal case?
2) are vmkernel and vmklinux one work or two separate works?

That 1) is nothing more than dirty legal play, easily countered by MORE developers involved standing up and being counted.

Surely, in this community, there must be more than one guy willing to stand up and be counted who has code that's affected?

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 17:55 UTC (Fri) by jhoblitt (subscriber, #77733) [Link] (3 responses)

The "only 1%" of source code argument seems disingenuous considering the scale of the work being considered. "1%" might be equivalent to man-years of effort.

VMware doesn't make it easy to find the vmklinux sources. I finally stumbled across a git repo the SFC has made available: https://git.sfconservancy.org/?p=vmkdrivers;a=summary

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 18:11 UTC (Fri) by jhoblitt (subscriber, #77733) [Link]

Running sloccount on the SFC repo puts the SLOC at 1,137,360 for ESXi-6.0-U1a.

Welte: Report from the VMware GPL court hearing

Posted Feb 26, 2016 19:32 UTC (Fri) by drag (guest, #31333) [Link] (1 responses)

> The "only 1%" of source code argument seems disingenuous considering the scale of the work being considered. "1%" might be equivalent to man-years of effort.

Don't try to apply conventional wisdom or logic to copyright. You just have to know the rules and know how they can get applied/interpreted by the courts.

Copyright isn't based on any sort of 'fundamental right of ownership', so there isn't much in the way of consistent basis. The purpose of copyright is to create a climate in the country's economy, by assigning restrictions on people's actions, that allows other people to profit from publishing their work. As such it's very arbitrary and there is little reason, besides international treaties, that one country's copyright rules match another's. It's all based on what type of rules are most likely to make the country wealthier.

Apparently in Germany as your 'work' is mixed in with other's it is possible for your claim on that code to be diluted. This is different then the USA's concept of 'derivative works', were claims seem to be pooled and do not lose their significance.. it just becomes more complex. There is the possibility that he will need the assistance of many different kernel maintainers to get enough power to successfully sue VMware in Germany.

However it seems that this is just preliminary BS they that they have to go through to get the courts up to speed on what is happening. Christoph owns a significant portion of the SCSI layer which ESXi depends heavily on.

Welte: Report from the VMware GPL court hearing

Posted Mar 1, 2016 23:46 UTC (Tue) by steffen780 (guest, #68142) [Link]

What you write is (I think, IANAL) correct under US law, but definitely not under German law. Under our law copyright isn't even called copyright, the literal translation is "creator right". It is considered a fundamental natural right. Which I think is preposterous, but hey.


Copyright © 2016, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds