Public domain software is particularly tricky situation for multi-nationals. Vetting an MIT licensed library is virtually a no-op, but some jurisdictions simply do not recognize PD licenses (let alone "do whatever the fuck you want" licenses people sometimes write while being clever). The one time I got stung by this it wasn't that hard to fix but things could have gone a lot different if we had more deeply leveraged either of the two PD tools that we had in our dependency chain. I swapped one for a slightly larger MIT licensed library, and set a coworker up to cleanroom the other, since I was already tainted.
I’m involved with some US government OSS projects and it’s interesting US government content is public domain in the US so not copyrighted. This seems like it should be easy for open source because public domain, no copyright can be used for pretty much anything.
However, this was confusing to many users and contributors so we worked with our lawyers to update the public domain notices to also include CC0 and MIT or ASL2.
That seems to help but we still get questions about why include CC0 and an OSI license. There’s some confusion that these conflict although they don’t of material is released under both consistently. I think there are issues if you try to add one on top of the other.
But I wish there was a cleaner way to handle this to allow for software and source to be used widely.
There is also an MIT-0 license that is OSI approved now. It doesn't have the questionable patent language that caused CC0 to eventually be withdrawn from OSI consideration.
Of possibly historical interest, the MIT license was created in the first place because there were issues with just putting X Windows into the public domain.
I don't see much problem with somebody saying "we don't recognize our property over X, but if you for some reason insist we have property over it, we allow usage under those conditions."
It's a really Kafkanian situation. And changing the license to make both claims not contradict each other would be way worse. But given the circumstances, it's a reasonable thing to write.
The USG cannot hold copyright in the US in certain cases, but it can still apply copyright outside the US. So one approach is to say "no copyright within the US, outside the US use this license".
The USG can hold copyright in some cases (e.g., copyright can be transferred to the USG). As often happens, "it's complicated".
There’s potential to invalidate the license because that first line is left blank or is inaccurate as sometimes people mistakenly add a copyright notice.
But our attorneys said that wasn’t a problem and we haven’t had any problems in the past decade, knock on wood.
The second paragraph of the Wikipedia article on USG copyright states this and cites a couple of sources:
This act only applies to U.S. domestic copyright as that is the extent of U.S. federal law. The U.S. government asserts that it can still hold the copyright to those works in other countries.[3][4]
Copyright for software makes it incredibly difficult to just fire and forget about a piece of code. I do not want attribution, I do not want to be held liable for it, I do not want to give support.
In the "default for-profit" view of copyright, not even wanting attribution is something that completely breaks the system and it saddens me.
It is the other way around. The concept of public domain only works in countries (eg. US) that have the common law for-profit copyright approach. In civil law jurisdictions where the copyright legislation more closely follows the idea of Berne convention the only way how something can enter "public domain" is by its copyright protection expiring.
> You could sue me for copyright infringement anyway, but you’d need to prove that my program copies yours. Some of the criteria used to prove this include the structure and organisation of the original source code, so even if I were to have used a different programming language, if I use the same code structure (even coincidentally) then it can appear as if I copied your program.
AFAIK, at least in the US, coincidence in and of itself is just fine, to prove infringement the copyright holder has to show that the defendant had access to the work, i.e. they actually copied it. This would seem to rule out most scenarios with a SASS/closed source plaintiff.
> US, coincidence in and of itself is just fine, to prove infringement the copyright holder has to show that the defendant had access to the work
That's right, which is why the Phoenix BIOS and others on IBM compatibles was never successfully challenged by IBM. The developers were given the API and were able to prove that they had never seen the implementation.
It is only for example one way compatible with say the GPL. GPL projects can integrate your work but may choose not to due to unfamiliarity but you may not integrate their work.
It also expressly reserves patent rights for yourself vs gpl 3 which disclaims them. If your software doesn't represent anything you intend to pursue patents on it may still put off companies who fear to be sued by anyone who buys it.
