A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
Posted May 19, 2016 21:23 UTC (Thu) by rahvin (guest, #16953)In reply to: A discussion on combining CDDL and GPL code by smoogen
Parent article: A discussion on combining CDDL and GPL code
You might be able to argue the spirit thing as part of a damages review or fair use type defense but IMO the courts are going to take the license and it's strict conditions pretty seriously. And IMO you want those strict conditions or as the article says, the bad actors take the loose view and suddenly the license isn't worth anything.
Posted May 19, 2016 23:52 UTC (Thu)
by smoogen (subscriber, #97)
[Link] (15 responses)
The core issue is how much the court system is going to treat the GPL like a contract and how much of it various other types of legal matters. The fact that it is going to be up to Oracle to decide on one side if they have any victims they want to get money from and on the other side kernel authors who will have to show they have enough standing is going to be a long complicated game. And in the meantime, I expect Canonical has done its math and decided that it isn't likely Oracle will sue, or that Linus is going to sue.. so they can get enough usage out of it to make a larger bit of the pie.. because frankly they know that a large majority of their users's don't care about licensing at all. So all upsides for them.
None of this is meant to be an endorsement. I don't like it, but I am not a lawyer and I have come to realize that law is so unlike how they teach you in the mundane classes that my best bet is to usually assume that whatever I think it should be like is really wrong.
Posted May 20, 2016 3:28 UTC (Fri)
by clopez (guest, #66009)
[Link] (1 responses)
What is beeing violated is the GPL, but only if you argue that ZoL is a derivated works of the Linux Kernel.
So, in my opinion (IANAL) the only who can claim damages to Canonical is some of the Kernel developers because of an infringement of the GPL derivative works terms.
But in order to do that they first have to demonstrate that ZoL is a derivative works of the Linux Kernel. And arguing that is going to be difficult given what the main author of the Linux Kernel (Torvalds) said previously about AFS: https://lkml.org/lkml/2003/12/3/228
Posted May 23, 2016 8:30 UTC (Mon)
by paulj (subscriber, #341)
[Link]
Posted May 20, 2016 22:05 UTC (Fri)
by giraffedata (guest, #1954)
[Link] (12 responses)
The paper simply states the conclusion that all western legal systems let courts follow the spirit instead of the letter of legal documents to prevent injustice - it doesn't back it up with anything and that is far different from the law that I know. The only time I've seen a court openly rewrite an unambiguous term of a contract is when the result of the literal reading would be absurd (which implies the people who wrote it would be shocked). I don't know copyright licenses, but I presume the bar is even higher there, because authors have a fundamental right to be selfish and withhold their work from anyone, so I can't see it being called injust for a license to fail to give someone permission to copy under whatever circumstance.
But the damages angle accomplishes the same thing. If you copy without permission, but the copyright holder isn't hurt by your doing so, you can't lose a lawsuit over it. If your copying accomplished the same goals of the copyright holder as if you had done what you had to to get permission, it would be hard for the copyright holder to win a lawsuit.
Posted May 20, 2016 22:08 UTC (Fri)
by mjg59 (subscriber, #23239)
[Link] (11 responses)
Is that true? Copyright infringement carries statutory damages in several jurisdictions.
Posted May 20, 2016 22:59 UTC (Fri)
by giraffedata (guest, #1954)
[Link] (5 responses)
I can't recall ever hearing about someone suing for statutory damages for copyright violation.
Posted May 21, 2016 18:22 UTC (Sat)
by nybble41 (subscriber, #55106)
[Link] (1 responses)
Statutory damages are basically the only kind a copyright holder *can* sue for, since there is no such thing as actual damage from making a copy, whether authorized or not. All those high-profile "piracy" cases where individuals were threatened with fines on the order of hundreds of thousands or even millions of dollars for copying a handful of songs or videos were founded on statutory damages. Of course, such cases require solid evidence and generate a lot of negative PR, so more recently they've been focusing on pressuring ISPs into serving as their extra-judicial enforcement wing and "persuading" individuals to settle rather than suffer the legal expense of defending themselves.
Posted May 22, 2016 3:08 UTC (Sun)
by giraffedata (guest, #1954)
[Link]
OK. That kind of copying isn't usually the kind that authors of open source code worry about, but it's certainly a valid example of statutory damages being real.
I'm not sure what you're trying to say by there being no such thing as actual damage from making a copy, but suffice it to say that people regularly win sums in court much larger than statutory damages to compensate them for the damage done to them by other people making copies.
Posted May 22, 2016 1:13 UTC (Sun)
by happylemur (subscriber, #95669)
[Link] (2 responses)
Posted May 22, 2016 3:03 UTC (Sun)
by giraffedata (guest, #1954)
[Link] (1 responses)
Posted May 23, 2016 11:01 UTC (Mon)
by paulj (subscriber, #341)
[Link]
You still need to persuade a court that there has been a copyright violation though, in distributing a derivative of CDDLed and GPLed works. I think the Eblen Moglen and Mishi Choudhary paper is more examining that question, than the damages issue, if I understand correctly (also, I don't think the paper is taking the firm position on the CDDL/GPL issue that this LWN article seems to paint; if Choudhary took a firm stance in the talk, the paper does not have it). That paper makes an interesting point on literal readings: under a very literal reading, one could not combine BSD or other code considered "GPL compatible" with GPL code, without explicit relicensing to GPL.
Interesting stuff.
Posted May 21, 2016 19:38 UTC (Sat)
by smoogen (subscriber, #97)
[Link] (4 responses)
Posted May 22, 2016 3:49 UTC (Sun)
by nybble41 (subscriber, #55106)
[Link] (3 responses)
Posted May 22, 2016 4:55 UTC (Sun)
by karath (subscriber, #19025)
[Link] (2 responses)
Posted May 23, 2016 2:09 UTC (Mon)
by nybble41 (subscriber, #55106)
[Link]
Posted May 26, 2016 10:27 UTC (Thu)
by Wol (subscriber, #4433)
[Link]
Likewise, in the UK, copyright infringement for gain (which is what Canonical are doing, and they are based in the British Isles) is also a criminal offense.
Cheers,
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
I don't even know why they brought up the "spirit" thing, because 1) that really isn't the law; and 2) the damages issue covers the matter just fine.
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
That's a good point. The paper doesn't mention that loophole. A lot of times statutory damages are just for measuring damages while liability itself requires that there be proof of nonzero damage. But I don't know.
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
Thanks. That sort of blows a hole in the idea that if the copyright holder isn't damaged then there is no liability and makes us rely for CDDL-GPL compatibility on the hypothetical power of a court to enforce the spirit of a copyright license, expanding the permission to copy beyond what the copyright holder actually said.
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
A discussion on combining CDDL and GPL code
Wol