[sb057]

It remains an open legal question (at least in the United States) whether or not an author even can voluntarily release something into the public domain. Hwaci could theoretically decide that, actually, that they never really put SQLite in the public domain and in fact possess full copyright over it and would like not only current and future licensing fees, but also retroactive fees from everyone who's used it over the past 20 years without legal permission. Lawsuits ensue, pandemonium happens, etc.

However, if you actually did acquire a license from them beforehand (when it was still theoretically free), they'd have no ground to sue and such a case brought against you would be thrown out of court immediately.

[cormacrelf]

Uh, no, they can't change history and take back the words they put there 20 years ago, or decide unilaterally that those words actually meant something else. The matter of whether a work has been successfully dedicated to the public domain (and when, and in what jurisdictions) is a matter for a court, not a matter to be arbitrarily decided or revoked at any time by the (at the time) copyright holder. It's a matter of words, interpretation, and copyright laws and treaties. Everyone has understood it to be public domain, this matches all the statements on all their documentation, and it would be ridiculous for a court to find otherwise.

The "open legal question" of whether you can voluntarily dedicate something to the public domain does not mean there is open slather to anyone who has done it to choose what the answer to the question is. It means it is open to a judge to decide the question once and for all someday, and then that question is resolved for everyone. I don't even know if your assessment of the question's unresolved status is correct.

[sb057]

>The matter of whether a work has been successfully dedicated to the public domain (and when, and in what jurisdictions) is a matter for a court, not a matter to be arbitrarily decided or revoked at any time by the (at the time) copyright holder.

Right, and a judge could rule that, in fact, it is possible to arbitrarily revoke free licensing at any time.

>it would be ridiculous for a court to find otherwise

You say that, but we live in a world of ridiculous IP law.

[cormacrelf]

> Right, and a judge could rule that, in fact, it is possible to arbitrarily revoke free licensing at any time.

On the basis of what law or treaty? First of all, dedication is not "free licensing". It is not licensing in any sense whatsoever. When you dedicate the work, you are no longer the holder, hence public domain is also what works fall into when the author is long dead. I'm not sure what piece of law you'd be re-interpreting to reverse all of that understanding, which would be a precondition to allowing dedicators to retain any control whatsoever over a work that's currently in the public domain.

Your original point was better, if it is ruled that "dedication to the public domain" is not and has never been a real thing, then the original dedications never took effect, and copyright holders remained copyright holders instead of forfeiting all control. Stick to that one.

[Beldin]

> When you dedicate the work, you are no longer the holder

I wonder how courts know how to distinguish between "I dedicate this work to..."

- "my wonderful husband"

- "my devoted readers"

- "the revolution in $COUNTRY"

- "the public domain".

Or perhaps there's no distinction and you all owe me copyright fees for those worjs dedicated to me as a devoted reader.

[cormacrelf]

Copyright assignment is what you’re describing, and yes they are able to distinguish, because dedication is not the word for that and more generally people aren’t stupid. Copyright assignment typically has a number of formal requirements to succeed including being in writing and being signed by the assignor. Next time tell the joke about a signed copy :)

[fanf2]

What prevents this scenario is estoppel: the users of sqlite have been told they can use it freely and Hwaci can’t retrospectively withdraw their promise. https://en.wikipedia.org/wiki/Estoppel

[riffraff]

every time I see this word as a non native english speaker it sounds to me as a low level error ESTOPPEL ("Error: Stopped by Eminent Lawyer" or something)

[marcan_42]

Even when the concept of "public domain" may not apply as such, one can create a license that effectively grants public domain rights. CC0 is an example of such a license. What SQLite is doing here is similar.

Jurisdictions may impose limits as to what rights can be waived via license, but those limits apply to all licenses. In other words, if a PD-style license is invalid and does not confer the right to use the software, allowing the author to sue you for using the software, then the same is true of all freeware, open source, and commercial licenses. Buying a license wouldn't magically grant you any additional rights just because you paid money.

The commercial license stuff is CYA for companies with legal departments that don't understand open source or think the words "public domain" are spooky. In practice, SQLite and anything under a similar license is basically "as free as the law lets us make it" - and if it turns out that's not very free, that means we have a bit problem with more popular licenses like BSD, MIT, and GPL.

In practice, AIUI, the finer points that are under debate depending on the jurisdiction are around things like retaining authorship and moral rights (i.e. being credited). I don't think the idea of being able to provide a piece of software for free with no restrictions on usage or modification is under any kind of serious question. And the idea of not requiring credit for derivative works is also universal in the entire copyright industry - when was the last time you saw a CD crediting the author of every single royalty-free sample used in its creation? So embedding SQLite into a piece of software is pretty uncontroversially fine.

Now if you took SQLite, changed all the licenses to say you wrote it, and tried to distribute it stand-alone like that, some jurisdictions may have a problem with that. That's where moral rights come in, and where "public domain" might not truly mean "public domain".

As long as you don't do that, you're fine.

[zarzavat]

I know that US courts do some weird things sometimes but "The author disclaims copyright" is abundantly clear, and also the author has made plenty of public statements clarifying exactly how that is to be interpreted.

The situation in civil law countries and especially countries that have inalienable "author's rights" is much less clear and hostile even to the SQLite copyright release.

[not2b]

But even in those countries, it would seem that only the authors, the people with "author's rights", have a claim on anyone's use of the code. If they intend for people to use and copy the code without restriction there is no one else to say otherwise.

[josefx]

> If they intend for people to use and copy the code without restriction there is no one else to say otherwise.

At that point you rely on a statement that is not legally binding and could be changed at any time. Wouldn't be the first time a software dev. rage quit and decided to go scorched earth on his projects.

[inopinatus]

Broadly speaking, licensing authority is not the problem. The absence of liability disclaimer is the problem.