The same happened to MySQL when Sun was bought by Oracle when MariaDB was forked - though I concede MariaDB isn't as popular as I hoped.
As it is, nginx is licensed under BSD - that would include the many contributions by anyone else (which cannot be claimed if the original project's copyright is found to belong to someone else) - so if the official nginx goes private but the BSD license is found valid, then the community can continue to publish their own "not-nginx" under BSD.
In a worse-case-scenario where copyright of the original nginx project are reassigned and the BSD license annulled, then the community need only replicate something resembling the original codebase that the community's subsequent (and fully legitimately BSD-licensed) patches/changes/commits can be applied to (though this will take some time) - so in either event, the prospective future "owners" of nginx cannot realistically hope to control the world of nginx deployments and the extended nginx developer ecosystem.
That has just been cast into doubt.
> copyright of the original nginx project are reassigned
Not reassigned. It would be found to never have been licensed.
> need only replicate something resembling the original codebase
Which is impossible without it becoming derived of the original. If you want to apply a patch to something it must first look like the original. A good chunk of the patches will have been derived from the original substantially. Which means they'd need to be rewritten too.
> the prospective future "owners" of nginx cannot realistically hope to control
Yes they could exert a lot of power and force people into paying. A clean-room effort would take a year! In that year they could go after everybody who is distributing old copies.
This is as bad as it gets if that company is found to own copyright.
I see this having only two outcomes: either nothing happens (it stays BSD), or it becomes nuclear waste, and new owner cannot sell it because it is plagued by mutually incompatible legal status of different portions of the code.
I can't see how the third version, where it stays clean, but the new owner can milk it, is possible.
The other patch authors could retroactively agree with the robber that their derived publications have been without license. Then they too could block usage of Nginx by pledging to go after anybody who gets a license from the robber.
If not, what's to stop other companies doing this in a predatory way? Start an "open source" project, gather years of valuable contributions from a enthusiastic community, then pull the rug?
Pulling the rug would mean revealing that your company never really owned the copyright, so at the very least you make yourself look bad.
Also, I don't imagine courts take a favourable view of transfers of ownership which are essentially fraudulent. I'm not a lawyer, as you can doubtless tell, but I presume it wouldn't be good for you if they could prove you'd planned the thing from the start.
There's an analogy outside of copyright: stolen goods. Sale of stolen goods isn't a legal transfer of ownership, but you don't want to be caught knowingly selling stolen goods, much less proven guilty of it (and the whole point of 'pulling the rug' is that things play out in court). I presume a similar principle would apply with intellectual property.
But since the project is under BSD license, they could make a case that they are allowed to use your changes (under that license). I'm not sure how that would eventually play out in court, though.
I would think that's more due to the huge boom in Postgres popularity than any failing of Maria. For the projects that do still use MySQL, all the ones I know of are on Maria or Aurora rather than the Oracle implementation.
Oracle has one goal, and that is to make money. They've realised due to the competitive pressure in this case that they can't abandon the community and go straight into value extraction mode.
It's all pretty arbitrary in my opinion -- if Monty and company hadn't whined so much when Oracle bought Sun, I doubt there would have been a fork.
Anyway... it never bothered me one way or the other.
Just wanted to note that it is not a very good way to gauge opressiveness or fairness... In a perfect world, charges would only ever be brought against the guilty, and only the guilty would be found as such by court. In such a World the ratio would be 0:<division by zero error>...
I agree on a perfect world scenario, but we in Russia have a big pile of ridiculous court decisions every month.
I would point out that the conviction ratio of the US is indicative of massive flaws in the powers that DAs have in most jurisdictions to force time limited plea agreements on defendants who have limited information and are under threat of charges with much higher penalties. This can usually be done without any oversight or approval from judges.
> In a perfect world, charges would only ever be brought against the guilty, and only the guilty would be found as such by court. In such a World the ratio would be 0:<division by zero error>...
That sounds like a world with complete surveillance, no privacy and a judicial system with absolute power. Doesn't sound like a perfect world to me.
It is usually 3 years for such cases, but it counts since the moment of time when alleged infringement was discovered.
Rambler used to be a somewhat successful and independent Internet company, but in dire straits recently because it's lost the market to its primary competitor - Yandex. So very much like Yahoo vs Google.
Now it's been bought out by Sberbank - the largest state bank run by a good friend of Putin. Sberbank is looking at becoming at becoming an IT monster, being a provider of just about everything - from food delivery to banking, with huge amount of Big Data at its heart. Delusions of grandeur at taxpayers' expense.
This lawsuit is the first big move of the new Sberbank's management team in Rambler.