> It's not that simple because the compatibility is from both ends; the CDDL states:
After reading this and thinking about it, I don't understand any argument a CDDL copyright holder would have against Linux? It frankly doesn't make any sense. You're going to have to explain the "Why?" of this. Start with -- "The facts are ... so the claim is ..."
> Linus in general is far more pragmatic about these sort of license issues than, say, the FSF or Stallman.
I wish they both, Linus and Stallman, would stop living in the 1980s. Both don't seem to understand modern copyright jurisprudence, from and including Computer Associates International, Inc. v. Altai, Inc.. I'm not sure it's intentional, but they certainly have both misled the public about what the GPLv2 entails, and their misinformation re: ZFS has been particularly egregious.
> Even if Oracle wasn't the Oracle we know and love today it would still be a risk: things can change in the future, companies get taken over.
AFAICT there is nothing special about Oracle as a litigant in this instance. AFAICT any Linux copyright holder would have have standing to bring suit.
> I don't understand any argument a CDDL copyright holder would have against Linux?
If you ship ZFS with Linux then those clauses I mentioned may apply. Whether that will hold up in court? I don't know. But do you want to run the risk of trying? I wouldn't, and I certainly can't blame Linux for not wanting to either.
I don't really know much about Computer Associates International. v. Altai, and I freely admit my ignorance on the finer points of US copyright law – it's not a topic I find especially interesting.
However, my point is that this doesn't really matter. Even if we assume you're 100% correct in this regard, that still doesn't mean Oracle can't and won't sue Linux. Anyone can sue anyone, and in the US the defendant is expected to carry their own legal costs regardless of the outcome of that suit unless the suit was filled under spectacular bad faith, which doesn't really apply here.
That is the issue; if ZFS was MIT or GPL then there obviously wouldn't an issue and any lawsuit would be completely baseless and any judge would throw it out in an instance. But with CDDL this is a lot less clear, even if we assume you're 100% correct on copyright, there still is a real dispute here with enough legal ambiguity that a judge will have to make ruling (i.e. they're unlikely to throw out the suit).
> AFAICT there is nothing special about Oracle as a litigant in this instance.
The Oracle-Google Java lawsuit stands out here. They argued that all the way to the Supreme Court with a novel and (IMO) creative view on copyright.
But like I said: it doesn't really matter. Even companies with good reputations can change, through change in management, change in owner, or just change of mind.
> If you ship ZFS with Linux then those clauses I mentioned may apply. Whether that will hold up in court? I don't know.
I'm sorry, but this is FUD by another name. This is a "Maybe there is something wrong with the CDDL too?" take, not a "This. See, this here is the problem with the CDDL" take.
FOSS people are supposed to be against this stuff.
> That is the issue; if ZFS was MIT or GPL then there obviously wouldn't an issue and any lawsuit would be completely baseless and any judge would throw it out in an instance. But with CDDL this is a lot less clear
Let me get this straight -- the law is complex, but it would be less complex if we had a different license, also we don't like this license and Oracle, so we won't even try with this license, even though it may not be an issue.
All of this is a giant turn off for me. I don't find it hard to believe that Canonical made the determination it did, because they're not beholden to a view that we should be scared of Oracle for indeterminate reasons, all the time.
And since you brought it up -- when Oracle sued Google, the basis of the suit was a claim that Google had violated the GPL because of, yes, another ridiculously broad interpretation of copyright law centered around the GPL. These over-broad interpretations, and attendant scared straight FUD campaigns, are not a service to the Linux/FOSS community, and Linux and the FSF and their minions should stop now. We should pick a side and the side should be in the majority of Oracle v. Google, and a long line of cases that say roughly similar things, not FSF, et. al., craziness.
After reading this and thinking about it, I don't understand any argument a CDDL copyright holder would have against Linux? It frankly doesn't make any sense. You're going to have to explain the "Why?" of this. Start with -- "The facts are ... so the claim is ..."
> Linus in general is far more pragmatic about these sort of license issues than, say, the FSF or Stallman.
I wish they both, Linus and Stallman, would stop living in the 1980s. Both don't seem to understand modern copyright jurisprudence, from and including Computer Associates International, Inc. v. Altai, Inc.. I'm not sure it's intentional, but they certainly have both misled the public about what the GPLv2 entails, and their misinformation re: ZFS has been particularly egregious.
> Even if Oracle wasn't the Oracle we know and love today it would still be a risk: things can change in the future, companies get taken over.
AFAICT there is nothing special about Oracle as a litigant in this instance. AFAICT any Linux copyright holder would have have standing to bring suit.