1. Software patents are a huge, huge, existential crisis facing our industry.
2. So huge, in fact, that nobody wants to deal with it. We can't realistically figure out if the libraries we are using are covered by patents, and given how broken issuance is, they probably are.
3. So the only practical solution is to use MIT licensed libraries, pretend there's no patent issue, and put our heads in the sand.
4. But Facebook's BSD+Patents license made us think about patents. Nobody stopped and said "can we use Angular? Are there patents on core elements of it?". But people did stop and ask that about React, because Facebook was (accidentally) making people think about patents, and the more you think about patents the more you wish you were a plumber and not a dev so all you'd have to deal with is toilets blocked with shit, and not software patents, which are much worse.
5. Now that it's under an MIT license it's just as dangerous to use React (maybe more) but we can pretend it's safe again.
Notice that Wordpress's decision to switch from React (which probably led to this) was quite clear: They were not switching due to patent concerns about React, they were switching because the non-standard license was stirring up FUD and making people ask them uncomfortable questions. (We know this because they explicitly said so, and also because the leading contender for a new library was Preact, which is strictly more dangerous from a patent point of view than React under every possible circumstance, but has a license that lets your pretend everything is okay easier.)
It's the backlash that was led by people who considered it a realistic possibility that they may, some day, want to sue Facebook for infringing on one of their precious patents.
This.
If the license is not one of the standard ones already in use, and it mentions patents, suddenly legal gets involved and it's a disaster. They're very risk-adverse and no one's really sure what will happen when it gets to court. It doesn't have to make sense - after all, most software patents don't make sense, and most software patent lawsuits don't make sense.
Copyright makes some sense - some entity wrote the code and has (or has not) given explicit permission for others to use it, and under common open-source licenses those terms are reasonably well understood. But patents can come out of left-field and be unrelated to the user or the producer of the code. Being aware of patents (which should not cover what you're doing or should not exist) makes punitive damages worse. Again, best to just put your head in the sand and hope the patents don't notice you. Any language about patents adds complications which no one wants to figure out.
Example: There is a patent in the US on ordering food from a menu online. It's held by a NPE (aka, a patent troll), and they charge a fairly high license fee. Some people who offer online ordering to restaurants pay the license fee; others don't. If you don't pay, then you can afford to charge lower prices than your competitors. If you do pay, then it's in your interests to tell everyone you meet about the patent, because every person you tell is someone who now faces triple damages if they opt to use on your your non-paying competitors.
Let me say that again, because it's so crazy: Some companies actually advertise their high prices, because a potential customer just reading their pricing page can thereafter no longer opt for a lower priced competitor without being subject to increased damages.
That's... sad.
> Being aware of patents ... makes punitive damages worse.
Sure, but that doesn't have anything to do with a BSD+Patents license. The license doesn't say "you also must do extensive patent research when using our software". You have to actually be aware of the patents that you're infringing to be held liable for extra damages.
Also, IANAL, so I may be entirely wrong.
But I read in other comments that some some lawyers think MIT implies a patent grant, which could explain why they switched from BSD to MIT.
Ahh, I see. But I think if they were just using BSD instead of BSD+Patents no one would have had any issue with it.
Facebook's patent clause disarms all their opponents, but leaves them free to attack. It's actually a wildly aggressive, offensive weapon, surprisingly - just like how nuclear missile defense is one of the more provocative technologies out there that possibly makes nuclear conflicts more probably.
That's why it's totally unpalatable, at least at the big companies.
"...if Facebook or any of its subsidiaries or corporate affiliates files a lawsuit alleging patent infringement against you in the first instance, and you respond by filing a patent infringement counterclaim in that lawsuit against that party that is unrelated to the Software, the license granted hereunder will not terminate under section (i) of this paragraph due to such counterclaim."
Moreover, nothing is preventing you from informing them of your patent and asking that they stop using it. If they continue to, it's clearly willful infringement.
Sure, not everyone is in that situation. Facebook is big enough though that if you have any patents you have to be _very_ careful about that license.
But on second reading I realize I was wrong (in a good way). Facebook's patent grant specifically says that if they sue you for patent infringement, and you counter-sue with something that is unrelated to the software of FB's that you are using, then your license to use FB's software is NOT terminated. That's actually entirely reasonable and great.
Patent grant terminated, not license
Oracle v. Google?
Using a license that's been in use for a long time, and has the endorsement of major OSS organizations, is pretty much always more comfortable for developers.
I wasn't uncomfortable enough with their licenses to rule out their projects based on the license, but I understand the folks who were. Even if my reading of the license makes me unafraid of it, there was enough uncertainty about how a court, particularly courts that have a history of siding with patent trolls, might interpret the license, for it to make some people nervous.
Also, the belief that the copyrights and patents involved will always belong to a benevolent actor ignores a lot of tech history. See: SCO, Oracle, etc. for examples of a previously decent stakeholder being consumed by a more malevolent and litigious one with a different interpretation of the law and the license. facebook may be a huge, unassailable, force today...can we be certain that in ten years that'll still be true? I'd bet on it, but I also won't be shocked if something dramatic happens. Sun looked pretty unassailable through the 80s and 90s, and then they didn't.
Big companies building big projects have to think about these things if they want to be responsible about the future of their projects. (Again, I agree that in this case, the fear is probably overblown. But, I get where it's coming from.)
As an aside, I realized I read FB's license wrong -- it actually is even better for users of it, and specifically says that the patent termination clause only goes into effect if the user is the aggressor. If FB comes after you for a patent claim, you can counter-sue all you want without losing your license. So FB is actually doing the absolutely correct thing here.
> With MIT nobody knows if you have those grants in the first place because there was never a court precedent. But not our hill to die on
I loved BSD+patent as one of the few potentially successful attacks on the awful software patents regime that we currently live under. This defeat means we are probably stuck with it for the rest of our careers.
Apache is an unethical foundation with shitty lawyers? I admire the goal of eradicating software patents, to me BSD+Patents was an unsuccessful way to do it. It hurt OSS projects more than it hurt private projects.
BSD + Patents doesn't look like the way we'll get there. Is there a way we can get there without putting "power in the hands of the big players" in the interim?
I get that there's legal uncertainty: I can imagine some companies, especially larger ones, not wanting to give up their patent-suit option just because one random small team in their org is using React. And barriers to adoption aren't a great idea when one of your goals is wide adoption.
Overall, though, I think making it much harder to file patent suits -- at least for software patents -- is a great thing. Facebook's (and others') implementation of it might have left something to be desired, and they're certainly a large company with a lot of resources behind them, so the power dynamic might be a bit lopsided, but...
Well, let's just say I'm a small company or individual who wants to release some open source code. Doing so under a BSD+Patents-style license gives me some protection from any big players who decide to use my software. How is this a bad thing?
I suppose I hold a pretty strong anti-patents stance, though. I'd be happy if (most?) patent protection just went away entirely.
As an aside, note that it looks like FB is going with straight MIT. So they're not even giving users of their software a patent grant at all anymore[1], and reserve the right to sue you later just for using their software, if they happen to have a patent on something in it.
[1] Yes, there are some legal opinions that the MIT license's language includes an implicit patent grant, but I don't believe that's a settled matter.