Hacker News Reader

The FSF and SFLC aren't "allowing wilfull violation"; they aren't the copyright owners of the Linux kernel, they just-so-happened to write some licensing terms (the GPLv2) which the Linux kernel authors decided to use of their own volition. If anyone's "allowing" something, it's the kernel developers.

Keep in mind that (a) the FSF uses copyright assignment for many projects, which would put it in a legal position to pursue infringements for those projects, but Linux is not one of those projects; (b) the SFLC is a legal organisation, not a software developer, so they only pursue infringements on behalf of client projects; as far as I'm aware, Linux is not one of their clients.

I've not waded too far into this before, but from reading http://softwarefreedom.org/resources/2016/linux-kernel-cddl.... it seems to me that:

- Canonical distributes Linux+ZFS binaries under the GPLv2.

- As per the GPLv2, Canonical must provide the corresponding source, also under the GPLv2.

- The ZFS source is under the CDDL, which permits relicensing binaries (hence it's compatible with the first point) but not relicensing of the source, which is incompatible with the second point.

- Hence Canonical can either provide Linux+ZFS source under the GPLv2, violating the source's CDDL; or provide Linux source under GPLv2 and ZFS source under CDDL, violating the binary's GPLv2; or stop distributing Linux+ZFS binaries. They chose the second option.

- Since CDDL and GPLv2 are both copyleft FOSS licenses, this incompatibility is arguably a technicality since it doesn't violate the "spirit" of either license, and may be ignored at the copyright owner's discretion.

- The copyright owners of Linux seem to be ignoring it, which is fine.

So I don't see what the big deal is. By acting in this way, Canonical improve the usability of their distro, whilst they and their users/customers incur the risk that some day, potentially, they might have to change their config to build the module separately (like Debian does).

And the only pertinent clause would be in the GPL (the CDDL doesn't care), regarding if ZFS were a derivative work of Linux, which it plainly is not.

All this "GPL and CDDL are incompatible!" is nonsense.

From reading that link, especially the Torvalds quote, it seems like all Linux kernel modules have always counted as derivative works, although Torvalds once told the AFS devs that he wasn't sure in their particular case, and he personally didn't think that theirs was a derivative work.

That seems to strongly contradict your "it plainly is not" claim.

As for GPLv2/CDDL incompatibility being "nonsense", I suggest you re-read the above. They are incompatible. The questions are:

- How much do the kernel devs care? They're the only ones in a position to complain, since as you say it's the GPL on their code being violated. The answer seems to be "not much", since CDDL is "in the spirit" of the GPL.

- How much would the legal system care if a complaint was made? As that SFLC link says, probably "not much", since it seems hard to argue that much/any physical or economic damage has been done, or any malicious intent (again, it preserves the "spirit")

- How much do Canonical care about the damage of a possible complaint? Clearly not enough to stop them going ahead.

- How much do users care about Canonical's risk-taking? Most users probably don't care.

I can imagine some 'enterprise level' (potential) customers who may run at the sight of "potential lawsuit", despite the "potential" seeming to be low, and the fallout from such a lawsuit being pretty insignificant (as SFLC says, since there's no malicious intent, it would probably just require a switch to Debian's model going forward, which would cause a little disruption but no real damage)

It is not nonsense, and Sun went out of its way to make sure that the CDDL and GPL are incompatible.

So, all the ZFS on Linux users are okay until Oracle sees money in suing them.

The link above seems to say that that the GPLv3 is compatible.

Linux won't(/can't) be changing from GPLv2 any time soon though. I don't know about Sun's intentions.

> Sun went out of its way to make sure that the CDDL and GPL are incompatible.

What evidence is there of this?

Sun could have written the CDDL to be GPL-compatible yet they chose not to and they generally refused to explain why. I suspect that they did not want to see OpenSolaris become an "organ donor" for Linux but they also didn't want to appear weak by saying so in public.

The closest I can find is where Simon Phipps said "to be GPL compatible you actually have to be willing to have your license discarded and replaced with the GPL. It's more a Borg-like assimilation than compatibility, in my view, and pretty undesirable as it results in future code enrichment being accumulated in some other code commons." https://web.archive.org/web/20050420234857/http://blogs.sun....

But is the Ubuntu ISO a derivative work of Linux?

The binary kernel in the ISO seems to certainly be a derivative work of Linux, but only the copyright owners whose code is included in that binary kernel would have standing to sue about this. That means parties like Linus Torvalds, Intel, LSI, and many other companies and individuals. And clearly so far those copyright holders haven't seen fit to sue about it.

It seems fair to anticipate no law suits since as far as I'm aware kernel developers have never filed a law suit against companies committing really blatant infringement. Most of the action there seemed to come from the BusyBox project.

First, things don't become legal precedent until someone sues and a judge rules on it (and preferably the ruling is upheld on appeal).

Second, what makes you think FSF/SLFC have standing to sue Canonical in this matter in the first place? Unless the rights of the FSF or SLFC themselves are infringed, they can't sue, even if they happen to have strong opinions on this matter. Either Oracle, some other OpenZFS contributor, or a Linux kernel contributor would have to be the ones bringing suit here - and of all these parties, probably only Oracle would actually want to do so.

The FSF/SFLC who are the authors and therefore guardians of the GPL maintain, publicly, that Canonical is in violation of its license. And yet, no legal action, no cease and desist has been sent.

In UK law, failure to enforce a claim or right is taken as reasonable assertion that the owner doesn't see the right as conscionable.

For better or worse, the FSF/SFLC is risking, through inaction which allows Canonical off the hook for claimed non-compliant behaviour, of having the soi-disant 'virality' of the GPL potentially ruled unconscionable (in Europe).

1) UK law should not be generalized worldwide.

2) The writer of the license has no special powers to interpret, enforce, or guard code licensed under that license. They can express strong opinions, which (if it's a well written license) garner attention, but they don't have any actual legal rights in the matter. Only the copyright holder (or someone they've delegated their rights to) can bring suit.

3) Even if the rights are estopped due to being left unenforced, if you go out tomorrow, contribute some code to the linux kernel, and then wait for canonical to link that with ZFS, you would almost certainly have your own right to bring suit. You're not estopped by someone else's failure to enforce their own rights, independent and unrelated to yours.

In any case, the arguments over ZFS have always boiled down to a few technical incompatibilities between the GPL and CDDL, mostly centering around some patent licensing clauses in the CDDL. While plenty of parties theoretically have standing to sue, in practice the Linux folks would love to merge it if they could, as would the OpenZFS authors - with the notable potential exception of Oracle, who remains silent on the issue. As such, none of the open-source developers are suing to enforce any sort of technical licensing problem because it's not in their interests to do so. And if Oracle gives up some rights, well, that's a good thing, right?

In short, most everyone involved doesn't want there to be a lawsuit - and so there isn't. And that's not a bad thing. The problem the FSF/SFLC keep going on about is that there's a lot of uncertainty - i.e. "someone _could_ cause you a lot of trouble by suing, so please stop doing this before you get yourself or your customers sued by someone else".

Just because someone uses your license, doesn't give you ownership rights or a legal cause of action regarding the code they licensed. What matters is who wrote OpenZFS, not who wrote the GPL.

Since ZFS isn't under the GPL I think copyright holder(s) of the Linux kernel, not the copyright holder(s) of OpenZFS, would have to sue Canonical.

The FSF and SFLC have as much authority to sue people for infringing the kernel's copyright as I have: absolutely none. Only the copyright holders can do that. Likewise, I can't sue people for trespassing on my neighbor's property.

In the US, (1) you can't let an issue get bigger simply so you can claim more damages the way SCO tried; and (2) when there's a specific statute of limitations -- like there is with copyright -- courts are less likely to say somebody delayed filing suit for too long, unless they made statements about trying to make a record to qualify for more damages, the way SCO did.

In the US, it used to be possible to lose copyright if you failed to enforce it, but that hasn't been the case for decades.

> The nature of law and precedent is that unless the FSF or SFLC sue Canonical Ltd. soonish for their distribution of OpenZFS Linux modules, their assertion that the GPL doesn't generally apply to kernel modules will become legal precedent by default.

Huh? Are you sure you aren't confusing copyright (which is automatic on creation of a work and requires no defense) with trademark (which requires a filing and which does require an affirmative defense)?

Now, you may lose certain damages because they passed out of statute of limitations because you didn't sue for copyright infringement if made aware of it. However, you don't lose the copyright and can sue for damages still in range and in the future.

Besides, in the case of ZFS, who is going to sue? Linus doesn't care. Canonical doesn't care. Illumos don't seem to care.

So, the only people who might care and have standing are Oracle, but bringing such a lawsuit won't get them money, so Larry doesn't care.

So, basically, Canonical can continue distributing it since nobody cares.

The SFLC (Software Freedom Law Center) does not think it's a violation. The SFC (Software Freedom Conservancy) does. It's easy to get the two mixed up. IIRC it was the SFLC's judgement that it wasn't a violation that made Canonical confident in shipping OpenZFS with Ubuntu.