Finally, people are beginning to realize the insanity of this entire PATENTS file situation!
When I first brought up how misguided people were for embracing React and projects with this license, I was downvoted to hell on HN. But really, everyone, THINK ABOUT IT. This is a company that glorifies and celebrates IP theft from others, and lionizes their employees who successfully clone others’ projects. They’ve built their entire business on the back of open source software that wasn’t ever encumbered with the sort of nonsense they’ve attached to their own projects. And this industry is just going to let them have it, because the stuff they are putting out is shiny and “convenient” and free?
Having known so many people involved with Facebook for so long, I have come up with a phrase to describe the cultural phenomenon I’ve witnessed among them – ladder kicking. Basically, people who get a leg up from others, and then do everything in their power to ensure nobody else manages to get there. No, it’s not “human nature” or “how it works.” Silicon Valley and the tech industry at large weren’t built by these sorts of people, and we need to be more active in preventing this mind-virus from spreading.
By the way, the fact that Facebook is using this on their mostly-derivative nonsense isn’t what should concern you. It’s that Google has decided, as a defensive measure, to copy Facebook’s move. Take a look at the code repo for Fuschia and you’ll see what I mean. Imagine if using Kubernetes meant you could never sue Google?
Yep, the patent grant is not fully reciprocal. As I understand it:
To use React, you give Facebook permission to use any of your software patents (edit: all patents, even on a new jet engine) throughout your whole company. Facebook gives you permission to use only the software patents necessary for React.
If your company sues Facebook over an unrelated software patent (edit: as above, any patent), your React license (to the React patents) is pulled and if your business has any dependency on React that part of your business is broken until you can port to something else, or you are in license violation if you keep using it.
On the other hand, if Facebook sues you over a software patent (edit: any patent) that isn't related to React, they are still free to do so (edit: you become free to defensively sue with other patents in response and not lose your license to React patents, meaning with the nuclear analogy below you give Facebook "first strike" capabilities but are allowed to keep a retaliatory-only arsenal).
So there is a big asymmetry: all of your software patents for a few of ours. (edit: all of your patents for a few react software patents)
It's a nuclear arms reduction agreement where Facebook gives up 1% of nukes in exchange for co-signers giving up 100%.
(edit: as pointed out below, it is a much worse deal: you give up all patents on everything, not just software patents; Facebook only gives up software patents on React)
> As I understand it: To use React, you give Facebook permission to use any of your software patents throughout your whole company.
That is, on its face, absolutely untrue. Nowhere in the React patent terms does it specify that you grant Facebook rights to use your patents--not even to ones related to React.
That's something that isn't in the React terms, nor does it seem to show up in any legal analyses published online from a lawyer making that claim. (None of the analyses I've seen, at least; if you found some, please post them.) In fact, the only place where this claim does crop up is in off-the-cuff message board comments.
FWIW, I don't use React, I don't want to, I'm not a Facebook employee, and in fact I think the world would be a lot better off with Facebook having less influence than they do today. But that doesn't change how weird it is to keep seeing comments like yours that make claims like the one above.
It maybe "on its face absolutely untrue" but it is true in practice. If Facebook infringes your patent and you sue them to enforce you lose the rights to use any patents React requires. So while you're not required to grant Facebook rights to use your patents, you are required not to enforce your rights relating to your patents against Facebook.
Companies shouldn't be dictating what happens if they're sued; it's a government right just like it's a natural right to accuse someone of theft. People shouldn't be sabotaged of their right to sue based on some arbitrary terms meant to block people from suing, their right to accuse someone of theft.
What if there was no patent grant, and just the BSD license?
Then you have no patent protection in the first place! You are in the same situation as when the patent grant expired because you sued them. I don't see how this is any better.
You do not give them a license to use any patent. You are right that in practice the non-enforcement means that they can willy-nilly use your patents. But if we are arguing about factual ability vs legal rights - they can probably already to that, due to the fact that they likely have more money and better lawyers then you do.
You are in a better place though - because you're not the only react user. If indeed it turns out there are some submarine react patents facebook owns, with this patent license most other users can continue to work. Without the license, facebook then has the choice to either sue everybody (kind of unlikely) or to create the situation wherein suddenly everyone knows that they're infringing but haven't yet been sued, or to retrospectively grant the license (probably losing trust).
And if you do have the license - then what? It's still a react death-knell should any submarine patent appear; no large company is every going to agree to such an assymetrical deal.
Better to at least be left in the situation where it's likely in facebook's own interest to never surface any (so restrictively licensed) patents in the first place.
> no large company is every going to agree to such an assymetrical deal.
The thing is, if you just use react for something inconsequential, and can live with the fact that you have to cease using it when you sue Facebook, then it is fine.
You shouldn't base your business on Facebook's (potential) patents, and then attack them with your own patents - but that is true with or without the patent grant.
Of course, what would be great would be 1) a finer grained, symmetrical patent grant, or 2) a bilateral agreement between your company and Facebook instead of the patent grant. If you are big enough and have patents, you can probably negotiate something like that.
> You shouldn't base your business on Facebook's (potential) patents, and then attack them with your own patents - but that is true with or without the patent grant. [...] If you are big enough and have patents, you can probably negotiate something like that.
The chilling future effect, which I believe is the primary driver of op worries, is what if you build on React and are the next Facebook? (I.e. you are growing quickly to become "big enough" but are not yet)
Old Facebook now how a lot more leverage on you at several critical moments -- either in terms of making an offer, or in the negotiations around a more reciprocal patent licensing deal, or the aforementioned infringing of your patents, or the mere threat of either if someone else makes an offer.
The entire point of ladder kicking is to build a moat to ensure that others can't grow past you in the same way you did, and this is pretty effective at that (intentionally or unintentionally).
As you said, a more reciprocal, more finely-scoped patent balance would resolve these concerns. (E.g. "if you sue Facebook over a patent involved in React...")
Yes, you're right that "BSD+patent"+"revoked patent grant because you sued Facebook over another patent" is pretty much the same as just BSD from the start. Licensing software you know you have patents on under BSD only with no patent grant would be pretty awful, of course.
If Facebook wanted to be the good guys here they would license under Apache 2.0 or write their own symmetric patent grant. They already relicensed RocksDB after Apache's move, perhaps they will do the rest of their stuff too.
> "BSD+patent"+"revoked patent grant because you sued Facebook over another patent" is pretty much the same as just BSD from the start.
Not really. If you have BSD with no patent grants you can at least argue that there's an implicit patent grant. If there's an explicit patent grant which you violated/got terminated that defence is clearly not going to fly.
My understanding is that if a company open sources a patent protected code (with one of the standard open source licenses without any additional patent clauses), the company grants anyone the right to use the patented technique without any obligations to the company. Otherwise we would see many cases of code released on liberal licenses, such as MIT, that no one is allowed to use, unless they pay for the patents separately.
That's definitely not the case. Some licenses like Apache-2 explicitly grant you the use of necessary patents. Others do not. And there are projects which are absolutely covered by patents, yet the code is opensource. That was the situation with Gif and MP3 formats for years.
I was involved with an opensource project which got a patent application submitted by the company. Even though I pointed out prior art. And even though I refused to sign the application as one of the authors. That's corporate environment for you.
As far as I know GIF and MP3 decoders were made open source not by the patent holders, if the patent holders open sourced the decoders, would they be able to enforce the patent rights?
The license of an open source implementation is granted by the authors and allows you to make copies and derivative works, while a patent grant comes from the patent holder and allows you to actually use the software. They are two completely independent things.
The fact that GPL licenses went for a patent grab (on hardware patents too) with an implicit patent grant paved the way for Facebook. It's just a very bad idea, unless you think all intellectual property should be swept into the dustbin of history, of course.
A company the size of Facebook can't really get away with that scheme even if it were intentional. For one thing, I can sell the patent to a troll and let them do my dirty work. Or a company could buy you solely to get fuel to sue Facebook over.
If Zuckerberg signs a paper with the wrong title, there's a troll lawsuit the next day. So if you make a habit of "in practice" strongarming people, you can easily slip up and have a huge headache on your hands.
> Nowhere in the React patent terms does it specify that you grant Facebook rights to use your patents--not even to ones related to React.
I detailed the specifics.
Take this example:
In exchange for doing graphic design on your menu, I stipulate that I can shut down your bakery if you take me to court when I eat from the bakery without paying. Now, sure, I haven't granted you the right to come eat my wares, but I'm going to have to shutdown my shop to keep it from happening.
In this case as long as you change the menu you can go back and enforce it, but with software, it isn't as feasible (say you had a game built on top of Raknet, which has the same PATENTS rider).
It's a de facto one-way "all your patents are available to us" grant, because you can't sue them if they violate any patents you may have, without ceding your l̶i̶c̶e̶n̶s̶e̶ ability to use React.
I'd call that a "distinction without a difference". Those are generally fallacious, and more often pedantic. Sure, you can still technically use React if, say, you're based somewhere that patents aren't particularly enforced. But if that's the case, why do you have patents in the first place?
That is plain wrong. Software patents exist and are enforced in the EU and other member states of the European Patent Treaty. The European Patent Office even has a nice FAQ on helping you get your software patents granted at:
The situation is a bit more subtle than you make out. The EPT has clear language which excludes software from patentability, but that hasn't stopped lawyers from arguing that some software can be patented if it has a physical effect (even a minimal one). Wikipedia actually summarises it decently: https://en.wikipedia.org/wiki/Software_patents_under_the_Eur...
The situation isn't even that subtle. The "physical effect" AKA the German Teaching has eroded into a "further technical effect" in the past decade. And with the UPC looming, things could get even worse. But that's getting off topic.
You can't legally use React without a patents license. The difference between the patents grant and the copyright license is mostly in the timeframe they cover (20 years for patents, 100+ years or whatever for copyright).
That nobody seems to know whether there are in fact any patents licensed by the patent grant within React is one of the most frustrating things about it.
Agreed. It gives the entire debate a very metaphysical quality. How can I determine the value of a patent grant when I don't know what the patents are, or if they even exist?
Now where does it say that? The way I read it, so long as you don't sue Facebook, you have a patent grant. If you believe people when they say that their patent portfolios are purely defensive, there is no problem.
On the other hand, being the aggressor in a patent lawsuit loses you the patent grant. This seems sane enough to me.
The normal state of affairs is completely ambiguous about patents. You get no explicit grant, so if they feel like suing, there's nothing to stop them. Some people claim an implicit grant, but there's no conclusive ruling on it.
I don't particularly care for software patents, but I don't see anything more stupid than the usual lawyerly CYA layered on an attempt to disambiguate the patent situation.
Why is it sane to lose the patent grant if you are a aggressor in a patent lawsuit against Facebook (for any patent, even ones unrelated to React)?
The clause also doesn't explicitly cover software patents only, so if Facebook infringes on your hardware patents (which is plausible, since Facebook does hardware related work as well), you still can't sue them.
I think the spirit is that patents are not legitimate. Suing for patent infringement is not considered exercising a legal right, it's considered being a dick and vandalizing someones property. So this is not to be read like a contract between to businesses, but like a truce.
It might as well say "we agree to not destroy your business using dirty tricks if you don't attempt the same".
That's how Apache 2.0 and GPLv3 treat patents. The language in the React license is not symmetric like the above licenses and thus doesn't feel like it's a "screw you" to patents. Not to mention that Facebook is very patent-heavy and it doesn't feel like they're anti-patent in any meaningful way.
I'd be more sympathetic to Facebook if they declared they would never use patents to sue someone and they'd only use their patents "defensively". Of course, they'd never do such a thing even though such a statement has zero weight.
> Suing for patent infringement is not considered exercising a legal right, it's considered being a dick and vandalizing someones property. So this is not to be read like a contract between to businesses, but like a truce.
This is an oversimplification. There are plenty of mum and dad inventors who would get destroyed by large companies but for relying on patent protection for their life's work.
Counterpoint - why would I want to build my business on software that, if the parent company screws me in some unrelated way, I have no recourse but to rebuild my entire software portfolio, or just accept that I've been screwed?
If they were offering something unique or especially distinctive, then maybe it would be worth risking, but as there are half a dozen or so alternatives, why bother putting myself into that particular pickle?
> Counterpoint - why would I want to build my business on software that, if the parent company screws me in some unrelated way, I have no recourse but to rebuild my entire software portfolio, or just accept that I've been screwed?
This doesn't prevent you from suing. It just prevents you from suing over patents.
Because they already allowed other people to use it, and it doesn't required extra effort on Facebook's part to allow another company to use it.
I can understand why Facebook would want to have a clause like this (to discourage lawsuits against them), but I would consider it against the spirit of open source to do so (ie. Facebook should be able to continue to do this, but I wouldn't consider React to be open source software in that case).
I wonder, does the definition of "open source" or "free software" require any patent grant at all?
I guess the GPLv3 has provisions for patents, but older licenses didn't. I remember some controversy around open source video codecs that you were not able to use safely because they were infringing patents.
Patents can be used to restrict all four of your freedoms under the FSF definition of "free software". They have written about this problem for quite a long time[1,2]. The problem with patents for things like formats or protocols is that they make free software implementations potentially dangerous to use (users can be sued for violating a patent). That's why h264[3], MPEG-whatever, FLV, etc are all considered to be bad even though they have free software implementations. MP3 recently shedded its patent licenses[4], and the response was that every distribution finally packaged mp3-lame in their main repos.
GPLv2 didn't have a patent grant because it wasn't a very well-known issue in 1991. Apache 2.0 was the first free software license to have a patent grant (which unfortunately made it GPLv2 incompatible) but GPLv3 included a similar (though stronger) patent grant because of Apache 2.0.
While these may sound like theoretical problems, people have been sued over using GIF, MP3, etc before. Every time you install Firefox, it will download a free software binary from a Cisco server that does h264 decoding[5]. However, they cannot legally recompile the binaries themselves as it means that the patent grant no longer applies and Mozilla (or your distribution) could be sued -- even though the binary's source is BSD-licensed.
> Now where does it say that? [...] If you believe people when they say that their patent portfolios are purely defensive, there is no problem.
Where does it say what? Which part of my post are you taking issue with, because I would love to correct anything.
What it certainly doesn't say is that Facebook's patent portfolio is non-aggressive. And when, outside of this, has Facebook said that about their entire portfolio (I'm not saying they haven't, I just haven't seen that before)?
My personal interpretation of this is that hypothetically you are likely to value your own patent(s) more than you value the time to re-write your application with something other than React. So if you need to sue Facebook, factor in that you’ll need a re-write as well.
My professional opinion is that if you use React and are concerned about this, talk to your attorney.
The thing with legal documents is that you can author whatever you please and none of it really means anything until it is brought to the courts to be enforced. At that point we learn what parts of what was written are actually valid.
Has any of this React patent usage agreement been tested yet?
But what that FAQ does not answer is: "Does the additional patent grant in the Facebook BSD+Patents license terminate if I sue Facebook because they violated one of my patents unrelated to the Facebook software?"
IANAL. One-sided contracts fail during legal process. There must be sufficient mutual benefit (mutuality in legalese). Has Facebook successfully defended the license's mutuality in court yet?
Maybe someone out there has access to legal cases and could help?
> To use React, you give Facebook permission to use any of your software patents throughout your whole company. Facebook gives you permission to use only the software patents necessary for React.
No, you lose the patent rights if you sue Facebook for any patents period.
> The license granted hereunder will terminate, automatically and without notice, if you (or any of your subsidiaries, corporate affiliates or agents) initiate directly or indirectly, or take a direct financial interest in, any Patent Assertion: (i) against Facebook or any of its subsidiaries or corporate affiliates
If you have to agree not to enforce any of your patents, that's pretty much equivalent to giving Facebook the rights to use them.
So if I understand correctly, if I were to use React for a marketing website about my patented system for <some extremely valuable invention> and Facebook steals/uses that extremely valuable invention, I can’t sue Facebook by virtue of using React?
Obviously, if that were to happen, you’d simply have to stop using React or else you’d get countersued for using React without a license right?
Why would anyone use React ever if that’s how it works? It feels like a payday loan – React makes it “easy” right now, but if your business is built dependent on React, you could pay a serious price later in exchange for “right now” convenience. It makes a strong case for Ember or Glimmer or any of the other front-end systems not tied to Facebook. Imagine a React “Native” app around which your business is based – you’d have to rewrite the entire thing of Facebook were to start using some patented aspect of your business.
Perhaps I am misinformed, but it seems absolutely bonkers to essentially indemnify Facebook from patent infringement just by virtue of using React somewhere in your business.
Imagine if Apple did this: if you release on the App Store using Xcode, you can’t sue us for any patent infringement ever or we’ll sue you for using Xcode illegally.
Is my assessment correct? Is React that good to risk such a situation? Facebook could sit back and watch a company using React develop something amazing, then Facebook could steal it and you’d have no recourse but to stop using React. Meanwhile, while you are rebuilding your application, you lose all your customers, but if you don’t rebuild your application, Facebook can legally take your invention and crush you in the market. Either way, you lose – simply because a few developers on the team were insistent on React’s greatness. If I were a VC, I’d be incredibly wary about supporting a company using React for their primary application.
Am I wrong? I am not trying to be argumentative – I just want to be sure I am understanding this correctly.
Yes, your understanding is largely correct. To be clear, you can still sue Facebook for stealing your patented invention, but then your grant to any patents that your use of React infringes would expire and Facebook could countersue over that. Facebook legal's response to this concern has been largely obfuscation and "trust us" which is not encouraging. They did change the patent grant two years ago to allow you to assert patent claims against Facebook without losing the grant if Facebook sues you first, but that's of limited utility.
It's funny you mention Apple, because the Apple Public Source License has a very similar provision.
> 12.1 Termination. This License and the rights granted hereunder will terminate:
> ...
>(c) automatically without notice from Apple if You, at any time during the term of this License, commence an action for patent infringement against Apple; provided that Apple did not first commence an action for patent infringement against You in that instance.
The good news is that the Apple Public Source License is basically dead and not used for anything of importance.
Please explain to me - are they able to change it again at any time in the future to "unclear" it again or just change wording altogether? Will it affect all users retroactively in that case? Because all this "they had bad wording in license but they changed it so we can use it now" sounds kinda crazy.
The patent grant is solely a grant from Facebook to the world, so Facebook can choose to grant more rights at any time. Just like I can sign a contract with you allowing you to borrow my bicycle on Thursdays and then later unilaterally decide that you can borrow it on Mondays as well.
The University of California did something like this in 1999 when they cancelled the advertising clause of the original BSD license. See ftp://ftp.cs.berkeley.edu/pub/4bsd/README.Impt.License.Change.
> You lose the patent rights if you sue Facebook for any patents that involve React
I see nothing that limits the termination clause to patents involving React. The only limitation following those general lines is the one you quoted, which is a limitation on the exception, implying that the license does terminate if the patent over which Facebook sued you is a React patent.
Is this widespread at all or is it just Facebook..?
Edit: I'm glad to be downvoted for asking a genuine fucking question if anyone else does this since Apache mentions Facebook and the only place I ran into PATENTS file is another Facebook property and googling brings nothing up except more React..
I love that while we discuss this awful attitude by the corporate giants like Facebook and Google, internally in "our own" (HN really isn't the community's, get a grip) we are CONSTANTLY worrying about getting downvoted for asking what we see as a genuine question.
There's a larger problem here, and I feel it comes down to our collective attitudes.
Important: IANAL and I am commenting from my personal reading
In the case of Fuschia, I don't see any more restriction of end-user rights than Apache 2.0 would give you. The only real difference is this line:
> This grant does not include claims that would be
infringed only as a consequence of further modification of this
implementation.
That makes it clear if you make a modification to Fuschia, that infringes on a patent that Google holds, that you don't automatically get rights to that patent too. Apache 2.0 doesn't mention that, but it appears implicit to me as Apache 2.0 only gives you patent rights to whatever is in the source.
Not to mention that patent law explcitly does not accept independent discovery as a defence for patent infringement, so I doubt that Apache 2.0 has any defences for that. I also don't believe GPLv3 does either but I would have to read through it again.
> They’ve built their entire business on the back of open source software that wasn’t ever encumbered with the sort of nonsense they’ve attached to their own projects.
So did Amazon apparently and others. At least Google have open sourced many interesting and useful projects so they have given back. Microsoft have been open sourcing things lately too. Others so so.
The problem is that a critical protection against large companies is via patents or at least the threat of. Without them, the ability to dethrone them becomes very limited.
Can you name an example of when a patent was used by a small company in the software industry to "dethrone" a very large company?
I'm having a hard time coming up with any examples; the lawsuits that come to mind are patent trolls, where the "little" company is a non-practicing entity so they can't be counter-sued for violating the giant's portfolio of patents. But, you sound pretty certain, so I want to figure out where I'm wrong.
I want an example of a case where a small software company successfully litigated a patent lawsuit against a huge company like Google, Microsoft, Facebook, etc. Preferably one where the instigator of the lawsuit was legitimately defending something novel and innovative. This is subjective, of course, and I tend to come down on the side of software not being something that should be patentable...but, it is patentable, so there are patents, so I'm asking: where are the lawsuits brought by small companies that are actually in the business they've patented and have sued a huge company?
You've made it sound like it happens all the time. I've been in the tech industry for two decades, and I can't think of any (though I'm sure they must exist). I assumed you had some cases in mind when you made the comment earlier, and I wondered what it was. I'm not demanding you go search for such a case. But, absent some sort of example, I'm gonna continue to assume it (almost) never happens.
Well, defending the company might entail getting acquired by another large company or NPA. I don't know of any examples where a small company sued a large one and continued to do business based on that lawsuit, but the threat of getting acquired certainly adds value to the company itself and discourages large companies from blatant copying.
Take that threat away, and small companies lose value and large companies copy without regret.
This is completely off topic from this thread, but one way to easily cut down on spam (that would break many email senders today) would be to require that a mail server can only send email for a domain you own. Yes, I'm familiar with DKIM and similar tools, but they don't go far enough.
Yah, I was particularly amused by Zuck when he gave a speech at his ol' stomping grounds suggesting everyone should focus on public works rather than companies competing with facebook. Very thematic, you have to give him credit.
If we can just remove emotion for a little bit, and coldly apply law and the terms of the license to fact, the picture is really not so dire.
- As stated in the standard BSD 3 clause, you're more or less licensed to do whatever you want with the software.
- The above doesn't give you a patent grant (at least not an explicit one), so Facebook also grants you a license to use any patents they have associated with the software.
- The ONLY possible way you can lose the above patent grant is if YOU sue Facebook for patent infringement.
Even if we assume, for the sake of argument, that everyone working at Facebook is a child of Satan, the only thing your company has to do to not lose the patent grant is NOT SUE Facebook for patent infringement. Seems pretty simple, no?
If your company is one that plans to, as part of its business plan, make money by suing companies like Facebook for patent infringement (which I have nothing against by the way), probably don't use react . Everyone else, what is there to lose?
Is the BSD-3 + PATENTS as nice as the Apache 2? No. Is it anything to get hysterical over? I can't see why.
It's because it covers any unrelated patents, not just the ones related to React. You can't sue Facebook or its subsidiaries, like Oculus or Instagram, for patent infringement, e.g. if Oculus were to infringe on your VR or AR patent, without losing a grant to any patents that may cover React owned by Facebook. It's unknown that Facebook has any patents for React, but if it did, it could countersue you simply because you rightly assert your patent against their infringing technology. If you own a company or are part of a university that has a patent and Facebook infringes, then you can't really use any of their open source products that have this specific grant, rather than not being able to use the specific product that they are infringing upon, if any. It's their prerogative to disallow that, but it goes against the spirit of open source, in my opinion.
I speculate that this grant was probably written in this way to ward off NPEs, but I doubt any of those NPEs actually would use React and so in reality it causes issues for pretty much anyone else that cares.
The real irony would be that if I found this video hosted anywhere else, it would be IP theft because the creators post it to Youtube. Your comment is unnecessary.
What percent of the top videos on Youtube were pirated from Viacom? Because according to the Kurzgesagt video, 73% of the top 1000 Facebook videos were pirated from Youtube.
Yeah, I've been trying since November 2015 to get Automattic to understand the Facebook BSD+PATENTS problem with their Calypso frontend to WordPress that uses React. Just added another comment to the issue linking back to this story:
"Replace React with Mithril for licensing reasons"
https://github.com/Automattic/wp-calypso/issues/650#issuecom...
Of course, that issue was rapidly closed with essentially a "we don't think there is anything to worry about here" -- ignoring that even if Automattic may not have patents to defend, any of their customers who has patents could face an issue using Calypso.
All this is made more confusing by the fact that Facebook changed the React license over time from Apache to BSD+WORSE_PATENTS and then to BSD+LESS_BAD_PATENTS. And it looks like Automattic may have started using React when it was Apache licensed. So they got the rug pulled out from under them by trusting Facebook. I can wonder if their code might be considered to have grandfathered Apache patent licensing rights?
The Fuchsia license by Google (at least the one file I looked at) does not seem to me to be that different from other FOSS Patent licenses in that it is not as broad as Facebook's and just covers patent claims about Fuchsia itself. So I would not put it in the same category as React's much broader PATENT retaliation clause which applies for any disagreement over any patents with Facebook whether they are related to React or not. So it is more like if you sue about Fuchsia patents you can't use Fuchsia.
https://github.com/fuchsia-mirror/magenta/blob/master/PATENT...
I don't think Google is doing the same - look at Polymer for example. You also have a patent grant but it expires if you try to attack possible patents related to Polymer itself - its like in Apache 2.0 which is more sane in my opinion.
If my company don't ever plan to sue Facebook for patents, is not involved with patents, doesn't care about patents and everyone leading the company thinks patents are bullshit, why should I worry?
You should worry because you don’t have a magic orb that lets you look into the future.
What if you got acquired. What if something unexpected happens. What if you got pulled into some suit against Facebook. There are a ton of things that could potentially happen.
To anyone concerned about React's virtual DOM and diff'ing stuff, and a potential Facebook patent thereof, in early 2012 I wrote and published (under a GLPv3 license) a virtual DOM implementation with efficient diff'ing when I forked SproutCore[0] to become Blossom.[1]
So even if Facebook tries to patent that particular invention/innovation, it may not stand up to legal scrutiny depending on the filing date. AFAIK, Facebook didn't do a provisional patent for virtual DOM stuff before July 2012 (long after I released Blossom), because that patent filing would have become public AT THE LATEST on January 1st, 2016 and nothing has come to light that I'm aware of.
So…you should be safe (IANAL).
[0] Ironically given the subsequent popularity of React, the SproutCore team rejected my virtual DOM approach which is why I had to fork it. Live and learn. I actually came up with the specific virtual DOM + diff design in spring 2008, but didn't get around to writing the code for it until someone paid me to do it (I had asked Apple and they declined). Eventually, the copyright owner of SproutCore (Strobe, Inc.) got bought by Facebook, though I don't recall when…
There are jurisdictions that don't recognize PD grants. I can't think of a good reason to use CC0 over MIT or BSD.
> GPL (any version)
This might be doctrinaire but I won't use the (L)GPLv3 because of the AGPL linking clause:
> Notwithstanding any other provision of this License, you have permission to link or combine any covered work with a work licensed under version 3 of the GNU Affero General Public License into a single combined work, and to convey the resulting work. The terms of this License will continue to apply to the part which is the covered work, but the special requirements of the GNU Affero General Public License, section 13, concerning interaction through a network will apply to the combination as such.
CC0, Unlicense etc. always contain fallbacks for jurisdictions without PD grants. That's the whole point of using them instead of just saying "public domain" and nothing more.
There's also the Zero-clause BSD License https://spdx.org/licenses/0BSD.html which is the same kind of thing except it does not even mention the words "public domain".
Still, be cautious. The OSI has approved a number of bad licenses. The UPL (https://opensource.org/licenses/UPL) stands out to me as one that should've been rejected.
The discussion around its approval was contentious. My reading is that it exists primarily as an Oracle CLA. The license wasn't rigorously reviewed by non-Oracle counsel and there's some worry about hidden landmines.
"I assumed the license was OK considering its popularity."
I would have too. I'm getting into Clojure(Script) where React(Native) seems to be a popular choice for UI. I haven't touched it yet. I hope there's a good alternative when I get around to it because I really like Clojure but I never wanted to rely on anything made by Facebook.
As far as I can tell the license is okay unless your product depends on suing Facebook while at the same time being so strongly dependent on React you can't switch it out.
I'm bewildered by the number of individuals who are stating surprise at how/why people use react despite it's license. It feels like a very out of touch perspective. What % of React users read licenses or need to care about licenses?
> use in source and binary forms, with or without modification, are permitted
You are licensed to use React even if you never agreed to whatever is in the PATENTS file. The reading where to use a source code is meant only to compile it, maybe admire the compiled artifact but never to put it anywhere near an executable environment seems pretty far-fetched to me.
Off the top of my head big users of React include:
* Apple (their entire dev docs site is built in it, and they've hosted Bay Area React user meet-ups)
* Twitter (for their mobile web)
* Microsoft (parts of Visual Studio use it)
* Netflix (all over their frontend stack)
...as well as Dropbox, AirBnb, Venmo...
So whilst I don't doubt you've discussed it with concerned counsel, there are huge corporations deploying React today, many of whom have extensive experience dealing with patent and IP issues (Apple, Microsoft, etc).
To add on to this list, even IBM with their large patent portfolio has a large number of teams deploying projects using React and other FB OSS technology.
Yes and for a while it wasn't allowed at IBM either for the same reasons. The latest changes made it possible again which is why now they can/do use it.
This is absolutely correct. Before the end of last year, getting approval to use it was a very challenging process.
However, many teams went through the process in order to use it because of how valuable they believe React is to their products. Groups like IBM Watson IoT were fortunate enough to get portfolio-level clearances as well during that time.
Thankfully the process changed and many product teams add React to their list of libraries/frameworks to evaluate for their use-cases.
To be concrete, IBM Watson has several products that leverage it that we sell to customers. For example, Watson Company Profiler, Watson Virtual Agent, Watson Discovery-related tools, and more.
Other organizations use it as well inside of IBM, like IBM Cloud and IBM Watson IoT, but I'm less familiar with their products off the top of my head.
I don't think so, I work for a big corporation, and our legal department contacted Facebook to confirm the interpretation and ask if a reworking of the license was possible.
It wasn't possible, consequently my company asked that we move away from React.
However, one thing that could be interesting to see is:
* do the companies using React are using the latest version or the last version without the patent clauses?
Big companies know they have the resources to devote to a code rewrite if it ever came to that. Small- to mid-size companies are the ones that have the greatest exposure.
Also, it's my opinion that the companies like AirBNB who use it are not considering that it will complicate acquisition by another party. They are basically taking themselves out of contention for purchase by a company that is in direct competition with FB. Maybe, evidently, they don't care but I just hope everyone using it is thinking about such things.
Small to mid-size software companies mostly don't have patents. Naturally they aren't going to be too worried about problems that will only affect them if they are wildly successful. They are more worried about becoming successful in the first place than they are about optimizing their level of success in the unlikely event that they do become that successful.
So, the OP had a question, which was ""Considering almost every single large tech company is using React, presumably there's not much reason for concern? IANAL. ""
I pointed out a number of these are concerned. There is in fact, reason for concern.
That doesn't mean people won't use it, it just means they may be willing to take the risk.
In fact, some of the counsel who are in fact, concerned, are on your list of big users!
If i have 100 billion dollars in the bank, i may be willing to take more risk around a random open source project than a small startup.
You don't seem to argue the original point at all (ie whether there is cause for concern), just point out "there are lots of people who use it who have extensive experience dealing with IP". I'm not sure how this means there should be no concern.
For example, i know at least a few of those corporations treat React very specially compared to most other open source (in how they use it, in how they try to segregate it, etc).
Vscode does not use react specifically for the licensing reason. I don’t know if they do now but when I was at MS this was the case. I know many teams at MS who evaluated react and lawyers strictly advised to stay away from it. I hear things have changed a bit, but you’ll never see Office and Windows use it.
That’s not “Office uses React”, it’s a “collection of robust components designed to make it quick and simple for you to create web experiences using the Office Design Language.”
Very limited “use”; mostly just enabling external developers to interface using React to their existing tools (that are not using React). If the Armageddon should come, and Microsoft is to enter a legal patent dispute/battle, stopping this React “use” is a no-op for Microsoft.
Microsoft builds Skype on React Native. They use React in production on several websites including outlook.com. They are using it internally and not just for external devs.
You'd be actually surprised how many teams inside Microsoft wrote their own smaller versions of OS libraries. It happened a lot less after Nadella era, but Ballmer era had a "we can't use anything not built here"
Funny enough that the author of Inferno now works for React team at Facebook. Inferno was a great project with perf that blows react out the water. Not sure if it’s maintained anymore.
If you're concerned about React, then Preact is actually significantly worse.
Consider: The risk for React is based on 1) speculation that Facebook has patents on core React concepts and 2) the fact that patent grant that comes with React terminates in some situations if you end up in conflict with Facebook. So for safety, it might be better to make sure you use code that Facebook has no patents on, which (if (1) above is true) means "code that is nothing like react at all".
A React clone that doesn't have the patent grant at all is vulnerable to all the same issues React itself is, plus more. It's strictly more dangerous than using React itself.
(Unless, of course, we assume there's some bit of React that's subject to a Facebook-owned patent that would not also apply to Preact. But having spent a few minutes thinking about it, nothing leaps out at me.)
> Its internals are fundamentally different and not a derivative work.
That would be a strong defence to a copyright claim. We're talking about patents, where there is no concept of "derivative works", and what matters isn't implementations but broad concepts.
> Software interfaces cannot be patented
Right, but what makes Preact concerning isn't that it offers the same API, but that it is built using the same underlying concepts.
Preact renders independent, encapsulated, nested, functional-ish components, usually but not always written in JSX, into a virtual DOM. Components are rerendered to the virtual DOM when their inputs change, and then diffing is used to efficiently sync the virtual DOM with the browser DOM.
If Facebook has some sort of patent covering the virtual DOM (although thankfully as far as I know, they don't), it's hard to see how it would not also cover Preact. And this is true even if, as I assume is the case, Preact shares 0 lines of code with React.
Edit: To be clear, I don't believe the patent situation is concerning, I think Preact is a great project, and I may well be using it in production soon. But if I was worried about Facebook's patent portfolio, I would be worried about using Preact.
Understandable, and agreed about my point being for copyright as opposed to patents. It's interesting to note that, since the majority of DOM rendering libraries now employ some variant of Virtual DOM (react, preact, mithril, glimmer/ember, angular, even polymer!), a patent claim covering the concept of Virtual DOM itself would likely be invalid - the presence of so many prior and contemporary implementations of that paradigm make would invalidate it on the grounds that it cannot be "Non-obvious" or "non-novel". My knowledge of this stuff is largely gleaned from Wikipedia and thus not to be trusted, but it seems like there would be little grounds for a patent suit over VDOM. Also interesting to note: there are few (if any?) other companies directly associated with any of the other VDOM implementations - can a patent claim even be made against an Open Source project?
Lots of large companies are not using React because of the licence. It's just that companies generally don't talk about legal matters like this openly. So you only hear about who is using it, not who isn't.
Often the large companies that are using React have patent cross-licensing agreements with Facebook (or other special arrangements) that make the issue moot for them.
Otherwise, getting your in-house council (or even your board, depending on the situation/risk) to approve of using React with the very broad patent restrictions can be an uphill battle indeed.
My most recent 'big' fortune 500 company put me under very strict legal terms regarding what type of software I was allowed to use. Any code I didn't write 100% myself had to be approved by legal before it could be used in software I wrote for them.
No, seriously, that is what the contract said.
In reality, the bulk of developers used whatever they wanted with no consideration. I assume other contractors were under similar legal terms, but not sure about employees. I was seen as a pain when I kept asking for formal permission for stuff.
>My most recent 'big' fortune 500 company put me under very strict legal terms regarding what type of software I was allowed to use. Any code I didn't write 100% myself had to be approved by legal before it could be used in software I wrote for them.
Back in the day, this was the norm.
>In reality, the bulk of developers used whatever they wanted with no consideration.
This wasn't - it was grounds for termination.
I remember stuff like - (for Windows C++) you can use Microsoft MFC and Microsoft STL - anything else you write from scratch.
but still many licenses have legal requirements. i.e. I was working at a fortune 50 rare open source project, we had an acceptable license list but still all the source code needed to be audited for dependencies licenses and the legal team produced the license file to be included in distribution with all the notices as required by mit and apache
Well, I guess it depends on what you mean 'almost every single large tech company', but React is used by Facebook (of course), Apple, Twitter, Yahoo, Netflix, Amazon, just to name a few.
I know atleast one of these companies (I can't speak for the rest) had a pre-existing bi-lateral patent agreement with Facebook which rendered the react patent clause irrelevant, allowing them to use react without concern. Without that existing agreement this would not have been the case.
you have only named a few though; 'almost every single large tech company' requires a particular definition of large tech company or a lot more than a few to be a useful(ly true) statement
...that's obviously not complete, either. If you think of a large tech company the odds are they are using react somewhere for something, be it public facing or internal tooling. Many of the companies that aren't actively using React do so because they have their own JS stacks to dogfood (like Google and Angular).
Well, anecdotally speaking of course, my company does work for most of the major (a) banks, (b) telcos and (c) insurance companies in Canada, and the majority of them are moving from Angular 1.x to React (because Angular 4 is a mess) for their front-end work, both on the public and secure side of things.
That's interesting. I've done a good deal of work in all three verticals and introducing a restricted license in any of their environments would have been a tough sell that involved a legal review and a well documented business case. I can think of a few where I can guarantee they would never allow it - and not based on any problems with React's license in particular, but because of internal history.
I wouldn't be surprised if the choice of front-end framework for website development is still below the radar of legal/compliance for some of these places..
A lot of these groups still see front-end tech as just HTML and CSS..
Are you joking, or you just don't know? Telcos alone have a huge list of patents which they are happy to use in the courtroom, and they have won cases in the millions USD. Verizon ranked 52nd in the world for patents granted in 2016.
Bank of America was ranked 124th for its number of U.S. patents granted in 2016.
wouldn't that mean that your company is covered by the react clause but your clients aren't? of course they're not worried, if they're a separate entity they aren't the one in the agreement.
I suspect that for a lot of those it's a case of "the development guys want to use this open source library" and nobody in legal has ever been asked. For large companies this is likely an annoyance move, with the argument that nobody who chose to use it is authorized to make that kind of patent - related agreements any more than they're authorized to sell office buildings or manufacturing facilities, but for small tech startups of the sort that Facebook once was this may be a huge hammer.
Thank you for telling us about preact. Is there a good react-native alternative? Our team has bound our codebase pretty tightly to react and react-native (optimized for code sharing).
I dunno "every single large company" seems like an overstatement. Maybe thats the reason some of big enterprises are still on Angular and Polymer - some go for Vue. Some of them also have BSD+Patents but its less broad.
Could I ask what those bugs we're and if you had filed an issue for them? I don't have any deployed projects based on it, but I've played with it and talked to the main dev a couple of times - he's usually very on top of it.
I'd be curious to hear about this too - I don't see any issues filed or commented on by him, perhaps the bugs are more perceptual mismatches with his React experience as opposed to actual bugs. We only recently released preact/debug, which brings with it some of the guardrails React ships by default - maybe without those people run into issues.
FWIW there are hundreds of companies (some very very large) using Preact in production. If there were showstopper bugs, they'd be fixed immediately.
I think this may be referring to "preact-compat" issues rather than Preact issues. I know the Inferno team recently had to change a lot of the state scheduling to be 1:1 with React 15/16 (i.e. batch sync schedule setState after events was very broken and there were issues with controlled components in Safari).
I'm not a lawyer, but perhaps Facebook's BSD+Patent license is not even open source.
It's tempting to consider the BSD license independent of the additional patent license. However, the OSI has not approved CC0 as being open source precisely because it expressly reserves patent rights [0]. In the OSI's justification, permissive MIT and BSD licenses may provide implicit patent license, so by themselves they are open source. However, like CC0, BSD+Patents expressly exclude this possibility. Indeed, Facebook's licensing FAQ deems the combined work of the BSD+patents to be the license [1]. Further, recent court case has shown that these licenses are not simply copyright or patent statements, but can be actual contracts [2].
Hence, we have to consider the BSD text + the patents file text as the combined license. This license is not symmetric and hence may violate OSI license standards. I've made this comment in the facebook bug report, https://github.com/facebook/react/issues/10191
There are two things here: The copyright license, and the patent grant. Copyright applies to the concrete implementation. You have to agree to the license to be subject to it, and to legally use the code.
A potential patent applies to any implementation. Even if you write a clean-room clone of React, if it uses the same patent, Facebook has a patent claim. But that means the patent grant is not specific to the code; it doesn't even require consent, Facebook could allow you conditional patent usage even without your knowledge! A corollary is that you are strictly better off with the patent grant, it imposes no additional constraints on you.
License with no patent grant: Facebook can sue you for infringing patents, even if you are using a clone!
License with patent grant: Facebook cannot sue you for infringing patents, unless you do it first.
----
Second, I think the philosophy behind the patent grant is twofold: 1) that software patents are not legitimate. Enforcing a patent is not seen as a legitimate right, but an annoyance, like pissing on someones lawn. From that point of view, it seems not asked too much from somebody to refrain from doing that. (I don't know if that was the idea of the people who drafted that license, but it wouldn't surprise me.)
----
Another, unrelated observation (and please don't invalidate the first observations if this one is wrong as internet commentators are wont to do):
I see nowhere in the license [1] that it requires you to take the patent grant. Is that true? It would be silly to refuse it, because you are strictly better off with it, of course.
> A potential patent applies to any implementation.
This is the point that most people seem to be missing, and it applies to much more than just React clones. If for Facebook has patents required for Virtual DOM, you are no better off if you use Vue.js or almost any other recent frontend framework. Just because you're using a project without the patents file doesn't mean those patents won't affect you.
Does anyone know what patents you are granted with the React license? Virtual DOM work that I and some colleagues did (and have patents for) pre-dates even the earliest internal Facebook experiments with the virtual DOM. We even spoke with the team at Facebook about our work well before React was even publicly released.
>A potential patent applies to any implementation. Even if you write a clean-room clone of React, if it uses the same patent, Facebook has a patent claim.
Sure, but that only rules out a reimplementation of React. The question most people face is "React or existing alternative X?". If your chosen alternative is still actively developed, it's probably notably different from React in concept and implementation (why else would it still exist).
Now if Facebook owns any react-related patents, you use all of them by using react, your alternative has a good chance to be sufficiently different to not be infringing.
In terms of your chances of using Facebook patents, any alternative is strictly better than React itself.
That's how I understand the patent grant too. Are people here really confusing a grant with some sort of restricting of their freedom or do we miss something? I frankly don't want to believe everyone, including apache foundation, believes this without getting their facts straight..?
The problem is that unlike the Apache 2.0 or GPLv3 patent grants, the React license restricts you from suing Facebook for _any_ patent (even non-software patents).
And note that that Facebook doesn't reciprocate that, and so if they decide to sue you for any alleged patent infringement then you cannot counter-sue for patent infringement they may be committing if you are using React. Also it makes you far weaker in negotiations about patent portfolios.
I personally dislike this whole patent business, but I can imagine even pro-patent people would also be against this.
In what way does it restrict you from suing Facebook?
From what I can tell the patent grant is only beneficial to users of the software, not facebook. And it exists to address the concerns people might have that, if they use any facebook tech, they might risk getting sued. If you decide to sue, you're as good off as before. But if you don't sue, you can be sure facebook will never sue you either for violation of the patents they granted.
> In what way does it restrict you from suing Facebook?
Economically, by losing the license to use React in your product once you serve them with papers. It's an implied gun to the head. Or poison pill, if you prefer.
You're not losing the license to use React if you sue Facebook. You only loose the patent grant if you do. In which case you're off as without a grant in the first place. At least that's how I understand it, so this still looks to me like people raging about something they simply don't understand.
I think there is a legal layer, and an ideological layer. From the legal layer, this is a non-issue. You are not in any way worse off than if the license was public domain or BSD (as far as I understand it).
From an ideological or appearances point of view, it seems a bit clumsy. People see it as a submarine to get access to their patents, rather than an insurance to prevent unnecessary patent litigation costs.
I wonder what other companies do with their patent grants, e.g. Microsoft. I remember MS had one that said, you may use our patent for the Ribbon UI, as long as you follow our UI guidelines and don't make an Office competitor (which I find pretty restrictive). But maybe there are other options out there that people would find more agreeable.
I think a unilateral patent grant for React in the license would be nice, but unrealistic. Maybe they should restrict the kinds of patent actions that trigger the invalidation of the grant, say only to certain patents, but it is hard to find a good compromise.
Actually, from a legal layer it also might be a tactic to weaken the bargining power of anyone who uses React (why would Facebook want to discuss a patent portfolio with a company that cannot sue them for any form of patent infringement without losing their React protection). They also would lose their effective ability to counter-sue in a patent case.
I don't like software patents on an ideological level, but even from a legal and practical perspective these license terms should be reason for concern. The first thing I read when I look at a free software project is the license, and when I first saw React a few years ago I stopped looking once I read the PATENTS file.
I was interested in React based on what I'd read and was figuring it'd be worth looking into, but this provides all the reason I need to avoid it - I don't forsee a situation where I would personally or as a small company be suing Facebook, but I could see developing something then selling/trying to sell it to a larger company. If my code comes with a big side of "oh, and if you buy this you won't be able to sue Facebook or its affiliated companies for patent infringement" that could significantly hurt sales chances.
I'll call that a false equivalency, because only in my wildest dreams can I consider anything I'm working on equivalent to Microsoft. Among the Microsoft-level things I lack I'll count their legal teams, patent portfolio, and negotiating clout.
Anything I as a solo or mostly-solo dev would put together to sell would almost certainly be sold to a company less than 0.1% Microsoft's size - even at the 1% level it's just as likely that a company would decide they'd rather just build their own version or get someone to do so as a work for hire.
Edit: I'm also mentally picturing scenarios where Facebook's terms might end up tested in court. A developer team using an open-source framework without understanding the legal ramifications is one thing, and a thing done by people not authorized to negotiate binding contracts for the company. A purchase of a company or significant software (not off-the-shelf) on the other hand is by definition done by someone with authority to negotiate binding contracts. It's not that hard to imagine a scenario where that difference becomes relevant in court. Heck, I wouldn't be surprised if it turned out that Facebook had an intern or two somewhere focused on identifying and tracking projects using React and what companies they're owned by, just in case they end up in a conflict with someone and need to say "Hey, are they using anything of ours that can make this go away?"
Well, my point was that Microsoft has a very broad patent portfolio, as does Apple, and many other such companies that are using React in production. If they do not fear this patent clause, then I fail to see why your startup would. Especially if your startup has no patents at all (as is the case with most software startups). Apple & Microsoft would have a lot more to lose if they were somehow granting Facebook unlimited access to their patent portfolio just by their usage of React. You think their attorneys would let that fly?
I don't think their attorneys would let that fly, but I also think their attorneys most likely have already negotiated other contracts with Facebook that override whatever's in the React license.
This is not a tool for Facebook to be able to use against another multibillion major tech company. This is a tool for Facebook to be able to use against potentially disruptive tech upstarts of the sort that they were 10-15 years ago. Also, don't just think about Facebook's core business - remember that they're getting into all sorts of fields now. Do you think there are any VR related patents held by smallish companies that might be of interest to the owner of Oculus? If those companies are using React, Facebook now gets to negotiate with them from a position of "You've already basically given us permission to use your patented IP in our new headsets."
You can't avoid react because it started a new paradigm. 9 out of 10 frameworks are based on the same principles and native follows suit where even big vendors are changing their strategies for mobile and desktop to align them with reacts principles (react-native, microsofts react-native-windows, reactxp and office fiber, googles kotlin for android, litho, componentkit, etc). Switching from react to something else is sometimes as simple as adding an alias in webpack. If you don't like the license, use preact, inferno, etc. But the idea itself is way too valuable to be in any way affected by petty licensing issues.
The paradigm is everywhere, so just use another library that uses that paradigm, like Inferno or Preact, or even something slightly different like Vue if you want.
The point is those other libraries could be in violation of any patents that Facebook may or may not have with regards to React. Perhaps they have a patent on virtual DOM? That would mean your usage of Inferno, Preact, Vue, or Angular 2+ would violate the patent.
That would frankly be impossible. All JSX boils down to is h(name, properties, children) and that is literally it. Each reconciler is different, uses another v-dom, etc. If you wanted to patent react, you patent the mere use of functions. And not just that there are about 20 years of fair use of similar principles, even for jsx in particular which evolved from mozillas e4x and otherwise keeps to basic xml markup and mustaches.
Also impossible because JSX is independent of React. The specification is only subject to a standard Copyright and is released under the Creative Commons Attribution 4.0 International license.
Well, I opened that issue quite some time ago, and opened it in response to a proposal by FB to use RocksDB in Cassandra, so it's a great change but not a single Fay turnaround
I worked or a large company on several web-based apps right around the time React came out. There were some UI issues I thought could be sorted easily using React.
After going to our lead dev, who in turn went to our project manager, we received an email from our legal department a few days later that simply stated we would not be using React due to "certain patent constraints."
Having not done any prior research, I looked into what the problem might be and was pretty floored with what I found. At first I scoffed when they said no, but after reading about the patent situation I totally get it.
Can someone explain how this can affect to projects using react, as in a part of a product of a company or personal projects? Thanks
I found this [1]:
> FB is not interested in pursuing the more draconian possibilities of the BSD+patents license. If that is true, there is actually very little difference between BSD+patents and the Apache license. As such, relicensing should make little if any pragmatic difference to Facebook.
So what happens if Facebook doesn't change the license and in the future changes its mind?
Facebook has reserved its rights with their license and they can decide to enforce it at any time. Personal projects should be ok, but commercial projects should involve a discussion with legal. If the company thinks it's possible that they end up in a lawsuit with FB, it won't be allowed. For those who get caught unawares, they will have to move off react if possible, or seriously consider settling.
For personal projects, my interest here is in whether or not the patents license is compatible with other free software licenses (say GPLV3). That might be a consideration for some people.
> My reading of that is if you use this code and Facebook infringes a completely unrelated patent that you own, you can't sue them for infringement without giving up this patent license. That is much broader than the language in the ALv2 which is limited to a single work.
So it almost certainly doesn't affect personal projects unless you are the sort of person who files for patents as a hobby, but for a commercial project, the scenario, I think, is that if you own some patent (not even a software patent!) and you sue Facebook for it, you lose your patent rights to React. In effect, you may as well be granting Facebook a patent license to literally all your patents for use in any context, in exchange for using React.
Also, it's not just Facebook: it also covers suing anyone for infringing your patent in their use of any other Facebook product or service, as well as "any party relating to" React -- so quite possibly, depending on how lawyers and judges read the license, you can't enforce your non-software patents against anyone who uses React.
I think this is misleading. In the event that you bring a patent infringement case against Facebook, you still have a license to use react, but you lose the patent rights that would cover you if you came up with something that infringed on a Facebook patent. Meaning that you now have to fight your case in court without any protection. You're safer with this license and the patent grant, than you are without it. The BSD license on it's own, and the MIT license don't give you any sort of patent grant at all.
That depends on whether you believe that the BSD license gives you an implied patent grant - it does, after all, just say "Redistribution and use...are permitted," without saying anything specific to copyright.
If the concept of an implicit patent license is coherent and legally meaningful (it sounds like it is), then an explicit patent license, like Facebook's, would supersede it - and so losing your rights under the explicit license would leave you worse off than relying on the implicit one.
In any case, this is why people like the Apache license and the GPLv3, so you don't have to think about whether a patent license is implied. Both of those have explicit patent licenses with significantly less draconian termination clauses than Facebook's.
(Obviously I am not a lawyer, otherwise I'd know better than to comment on HN about unclear areas of IP law.)
The MIT license almost definitely protects you from anyone who published under that license and would try to bring a patent suit against you. It enumerates protection against anything a suit might deal with. It just does it without actually using the word "patent" anywhere.
Not necessarily. Case in point: Its language regarding patents is what made the OSI reject CC0 (or discuss it to death until its application was withdrawn):
Without an explicit mention of patents, it's up to the courts to decide whether the grant of usage rights includes an implicit grant of patent right. With an explicit mention, you're hosed.
> In effect, you may as well be granting Facebook a patent license to literally all your patents for use in any context, in exchange for using React.
I'd say that's overstating the case. You aren't granting them a license to your patents and you can still sue them for infringing (which you would not be able to do if you granted them a license).
Rather, you are promising to look the other way for as long as you are using React. Should you choose to sue,
1. You can say in your case that they never had a right to your patents
2. You can take the time to remove your usage of React prior to filing the court case (so that you never infringe).
Point 2. sounds kinda sketchy. Facebook's BSD says that even 'take a direct financial interest in' suing is forbidden. So if you begin cutting React out of your operations and then sue them already React-free they will probably try to claim that for a time you were using their patents in React illegally because you considered suing them before removing it and at that moment the patent grant terminated.
Glad the long-term legal implications will be given serious consideration publicly, rather than the "this is not the droid you're looking for" I've seen nearly everywhere so far!
Last update is interesting. Someone did a search and looked for any patents FB filed for and couldn't find any in last year. So somehow based on that they decided things are "OK".
To quote:
---
US allows to patent up to a year after the publication, allowing to conclude that ZSTD remains free of patents (?) - suggesting this "The license granted hereunder will terminate, automatically and without notice (...)" from PATENTS file has no legal meaning (?)
---
Anyone care to validate how true that statement is?
They do not have to publish. The USPTO will publish patents (provisional or full) 18 months after filing, so -- at the very least -- one has to wait 18 months.
It is also quite possible that facebook collects patents under a different name, e.g. "faceless intellect llc", a wholly owned subsidiary of Facebook. Or they may have bought patents that are essential for ZSTD from someone else who filed them independently.
The conclusion that "zstd remains free of patents" is baseless and made by someone who is not familiar with patent practice.
> They do not have to publish. The USPTO will publish patents (provisional or full) 18 months after filing, so -- at the very least -- one has to wait 18 months.
Holy shit, what?! The US patent system means that you can infringe on patents that you had no way to know even existed at that point in time? How the fuck can someone look at this policy and think that it's in any way reasonable?!
They also publish at grant of course. It's just that it rarely takes less than 2 years from filing to grant so first publication by USPTO is usually pre-approval, 18 month from filing
IANAL but I have a couple of software patents (now assigned to Open Invention Network).
The line "A year has passed, but searching patents with 'facebook compression' suggests no patent application regarding ZSTD has been filled" is not very convincing to me for a few reasons, primarily that the word "compression" may not show up at all in the patent document. When I got the text of the patent application back from the lawyers, after describing to them in engineering terms the thing we'd invented, I could hardly understand what it was talking about. Patentese is its own dialect of English, and there's no reason to assume that it'll say in straightforward terms that it's a compression algorithm.
And I'd find it highly unlikely it contains any patents too. Occulus was bought in March 2014 by Facebook and RakNet was bought in April 2014 by Occulus and there is only a handful of commits on GitHub that don't seem substantial at all and are just maintenance and cleaning up the repo. RakNet operated for years before that so it's unlikely they happen to have had a patent in it at the time of its purchase or they'd be nuked by Facebook before.
Then again, there are/were patents for some bit tricks and Carmack's Reverse so anything is possible.
In the discussion they say RocksDB will be relicensed under dual license Apache 2 and GPL 2.
There's already an issue [0] asking them to consider doing something similar for react, and Dan Abramov said he'd route the request internally on the next work day.
I can't imagine they'd keep the existing license without harming their community image. But even if they keep the license, many applications should be able to easily migrate to preact [1] and preact-compat, which provides a react-compatible API.
Hopefully they relicense the project. It seems like it's the first thing that gets brought up each time react gets mentioned.
A React developer is on record some months ago as saying that as far as he personally is aware, no such patents exist. Further, patents are a matter of public record, and to date, nobody has identified a patent that might cover React.
Of course, a random comment from a React dev on github carries no legal weight, and absence of evidence is not evidence of absence. But so far the picture being painted seems consistent with there being no list of patents at all, secret or otherwise.
Then it will be relicened under "Apache 2+Patents" and "GPL 2+Patents"
All licenses to "use" and "distribute" this software are now contingent upon a patent. The copyright license is secondary to an overriding control and use and distribution via patent.
If Facebook has patents that cover React functionality. They almost certainly cover parts of other JavaScript frameworks. React is well executed, but it's conceptually simple.
I don't think avoiding React makes you any safer. You don't know how broadly Facebook or the courts will interpret their patents.
The license to use React is governed by the BSD license, and it's not terminated if you end up suing Facebook over their violation of you VR patents or whatever.
Edit: Downvoted for a factual answer? There's a host of sources confirming by answer as correct, including Facebook's official FAQ on React licensing[0].
Q: "Does termination of the additional patent grant in the Facebook BSD+Patents license cause the copyright license to also terminate?"
> I mean, if you can use the software without the patent license, what is the patent license giving you?
Facebook may or may not have some patents covering concepts or algorithms used in React. If they do (nobody is aware of any), then they could sue you for using React, or any other framework which uses the same underlying concepts (eg, Preact, maybe Vue, possibly Angular 2.x, etc., depending on what the patents cover). If they did sue you, a valid patent grant (which you obtain by using React) would be a very strong defence, and would be quite useful.
I'm not entirely sure what you're trying to say. The patent grant applies to Facebook's patents (if any) on the tech used in React.
> I think the patent grant applies to everything you make
Yes, if you have a patent grant from using React, it would be a defence from claims that things you have written that do not use React violate patents included in the grant.
I just re-read it, you're correct. It only covers React
> For avoidance of doubt, no license is granted under
Facebook's rights in any patent claims that are infringed by (i) modifications
to the Software made by you or any third party or (ii) the Software in
combination with any software or other technology.
>The license granted hereunder will terminate, automatically and without notice,
if you (or any of your subsidiaries, corporate affiliates or agents) initiate
directly or indirectly, or take a direct financial interest in, any Patent
Assertion: (i) against Facebook or any of its subsidiaries or corporate
affiliates, (ii) against any party if such Patent ?> > Assertion arises in whole or
in part from any software, technology, product or service of Facebook or any of
its subsidiaries or corporate affiliates, or (iii) against any party relating
to the Software.
Its not talking about terminating the patent grant, its talking about terminating the license to use.
The section you quite is under this one:
>Facebook, Inc. ("Facebook") hereby grants to each recipient of the Software ("you") a perpetual, worldwide, royalty-free, non-exclusive, irrevocable (subject to the termination provision below) license under any Necessary Claims, to make, have made, use, sell, offer to sell, import, and otherwise transfer the Software. For avoidance of doubt, no license is granted under Facebook's rights in any patent claims that are infringed by (i) modifications to the Software made by you or any third party or (ii) the Software in combination with any software or other technology.
Emphasis mine. So it seems to terminate just the patent grant. You can't terminate a BSD license so if you're in like France (no software patents) then you are fine to use React and sue Facebook for patent infringement.
> Its not talking about terminating the patent grant, its talking about terminating the license to use.
This is totally incorrect, and has been debunked many time.
"The license granted herein" is the patent grant, because that language occurs in the patent grant. The BSD license is totally separate.
This has been stated, repeatedly, by representatives of Facebook and independent lawyers; the language is standard and unambiguous. They even answered it in a FAQ here[0]:
Q: Does termination of the additional patent grant in the Facebook BSD+Patents license cause the copyright license to also terminate?
INAL - But this seems strange to me. The Apache license has what I see as being the same patent grant, with the same condition that if you make a claim against them, you lose the patent grant.
Apache 2.0
> 3. Grant of Patent License. Subject to the terms and conditions of this License, each Contributor hereby grants to You a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable (except as stated in this section) patent license to make, have made, use, offer to sell, sell, import, and otherwise transfer the Work, where such license applies only to those patent claims licensable by such Contributor that are necessarily infringed by their Contribution(s) alone or by combination of their Contribution(s) with the Work to which such Contribution(s) was submitted. If You institute patent litigation against any entity (including a cross-claim or counterclaim in a lawsuit) alleging that the Work or a Contribution incorporated within the Work constitutes direct or contributory patent infringement, then any patent licenses granted to You under this License for that Work shall terminate as of the date such litigation is filed.
The important bit being...
> If You institute patent litigation against any entity (including a cross-claim or counterclaim in a lawsuit) alleging that the Work or a Contribution incorporated within the Work constitutes direct or contributory patent infringement, then any patent licenses granted to You under this License for that Work shall terminate as of the date such litigation is filed.
But what people seem to be missing (as far as I can tell) is that you don't lose the licence to use the software. You just lose the patent grants. But with the BSD licence alone, you lose both the patent grand AND the licence. I really don't see how the Apache 2.0 License and Facebook's BSD+Patent Grant are any different.
The difference you're missing is that the Apache 2.0 terminates if you end up in a patent dispute related to the software. Facebook's parent grant terminates if you sue them over any patent.
So hypothetically if Google released some JS frontend library under Apache 2.0, I don't lose the patent grant if I sue them for violating my company's VR patents, whereas I do lose my patent grant to React if I sue Facebook for violating my VR patents.
My company happily uses React, because we don't believe that's a risk worth worrying about, but it is a difference, and it might be relevant to some.
I'm neither and English isn't my first language (not that legalese is even close to any sane kind of English from last 200 years..) but due to my deep interest in Free and Open software, old games reimplementations, abandonware, reverse engineers and emulation I spent quite a while reading about legal stuff.
The part you said is important sounds like you lose the right to use the patent in the given software if you attempt to sue someone over patents in that software which sounds absolutely sane.
So basically the difference boils down to the CAPS part below:
Apache 2.0: we grant you use of these patents, sue anyone over one of the patents we gave you and we take that grant away.
Facebook's BSD: we grant you use of these patents, sue anyone over one of the patents we gave you OR SUE US OVER ANY PATENT AT ALL and we take that grant away.
Another difference is what is covered. Apache is about suing, Facebook's BSD says you can't even 'take a direct financial interest in' any of the stuff they don't want you doing.
Another difference is what is covered. Apache is about suing, Facebook's BSD says you can't even 'take a direct financial interest in' any of the stuff they don't want you doing.
I do fully support this bit though, which is specifically to avoid that whole thing where Microsoft was bankrolling patent trolls.
This says you can't even seriously financially consider suing them or even call them out in any way over ANY of your patents.
If you happen to be a ML/VR/cars company and think "oh, we use React on one tiny website that'd be cheap to rewrite and Facebook uses without paying our 42314 super expensive patents for ML and self driving cars and VR.. should we sue them?" you already are using React patents illegally because you took financial interest in patent assertion.
With that wording taking out the "you (or any of your subsidiaries, corporate affiliates or agents)" part would still cover cases like bankrolling patent trolls that sue.
And 'patent assertion' means:
>A "Patent Assertion" is any lawsuit or other action alleging direct, indirect, or contributory infringement or inducement to infringe any patent, including a cross-claim or counterclaim.
So basically, thanks to that language, being that kind of a company I mentioned above, you could write a blog post saying "WTF, we use React and Facebook uses our expensive shit for free!?!" or even a legal letter to them to stop using patents without getting a grant from you and lose your patent grant already because you are 'alleging' or 'taking financial interest' in 'alleging' or 'lawsuit'.
And other users of software are only protected from you suing them over React patents like in Apache license, Facebook is protected in a special separate way from any patent lawsuit, allegation or even financial consideration of a lawsuit or allegation.
This is very one sided and the reason why I immediately remembered RakNet PATENTS file I've seen back in 2014 when I got hyped over RakNet being open source is because it stuck in my mind how stupid and one sided that patent grant was.
I don't understand how code with a bsd license without a patent grant is better for the apache foundation than facebook's bsd + patent license. With the former, the entity donating the source can sue you for patent infringement at any time.
Clearly the apache 2 license would be preferable (and what rocks db did), but I am puzzled the foundation accepts bsd code in their products, given their worry about patents.
How did Facebook was able to change the license of RocksDB so easily? The CLA is not a copyright assignment and therefore all contributors have to agree for the change. Did they contact anyone who has signed up the CLA?
2. Grant of Copyright License. Subject to the terms and conditions of this
Agreement, You hereby grant to Facebook and to recipients of software
distributed by Facebook a perpetual, worldwide, non-exclusive, no-charge,
royalty-free, irrevocable copyright license to reproduce, prepare derivative
works of, publicly display, publicly perform, sublicense, and distribute Your
Contributions and such derivative works.
This is not a copyright assignment and not an authorization to publish under a different license. Canonical's CLA adds the additional paragraph to enable that:
Based on the grant of rights in Sections 2.1 and 2.2, if We
include Your Contribution in a Material, We may license the
Contribution under any license, including copyleft,
permissive, commercial, or proprietary licenses. As a
condition on the exercise of this right, We agree to also
license the Contribution under the terms of the license or
licenses which We are using for the Material on the
Submission Date.
I hate to say it, but you have a perfectly valid point.
I hate to say it because it makes changing the license significantly more problematic than it seemingly was an hour or two ago. By changing the license, Facebook assumed rights to contributions that it does not have. Even if the licenses are _compatible_, releasing contributions under a different license is not allowed without permission from the contributors.
This is a change from a permissive open source license to another permissive open source license, but what if my GPL contributions to a project were released under BSD? Or what if they were released under a proprietary commercial license? In either case, I might not have contributed, and I should have the right to rescind my contribution to prevent it from being included under a new license.
It's an interesting quagmire, and one that FB's choice of license was destined to lead into at some point. I think it was a bad decision from the get-go.
This is not the first major licensing snafu. A lot of projects ran into similar issues trying to adopt GPLv3 when it first came about. And I have seen a few other projects deal with community uprisings as other license changes came about.
Personally, I think FB should adopt the combined apache/GPL license, adopt CLA language like you've stated, and provide notice that developers can rescind contributions if they don't agree. Perhaps do a tri-license of Apache/GPL/BSD if there's significant kickback. Their current license is a problem and too much of a continuing distraction. JMO.
I'm splitting hairs, but 'sublicense' and 'sublicense under any license of our choosing' are two different things. I'm surprised FB's CLA isn't more explicit.
True, it's not a copyright assignment, so Facebook should probably maintain the list of copyright owners somewhere (e.g. Go has (C) Go Authors and the list of authors is in a separate file), however, the CLA terms I quoted allow sublicensing. When signing the CLA, you grant Facebook the copyright license to your contributions, which Facebook then sublicenses to everyone else with the license of their choosing (as long as this license doesn't grant more rights than in the CLA.)
I honestly don't know what "sublicensing" exactly means. However, Canonical's CLA which explicitely says they can relicense at will has a special clause for that (and as a result, many people don't want to sign this CLA). I don't think that "sublicensing" (which is present in many CLA people are happy to sign) means the same thing.
Sublicensing means granting some or all rights licensed to you to third parties. The original license agreement for contributions is the CLA, the receiving party of the CLA sublicenses some of the rights from the CLA to others in the form of the project's copyright license.
On purely copyright grounds, any BSD licensed code can be relicensed as Apache 2.0 code. And the patent grant was solely from Facebook so it can be modified at their discretion.
This is incorrect. Here's the full text of BSD license:
Copyright (c) 2011 The LevelDB Authors. All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
* Neither the name of Google Inc. nor the names of its
contributors may be used to endorse or promote products derived from
this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
View
MIT license includes the right to sublicense, however the terms of sublicensing is retaining the original copyright and permission notices, so you'll still have to retain the large portion of the MIT license text.
First paragraph of BSD license says: "Redistributions of source code must retain the above copyright-notice, this list of conditions and the following disclaimer".
I agree you can add the Apache 2.0 license on top of that, but you cannot discard the BSD license unless you own the code.
> The license granted hereunder will terminate, automatically and without notice,
if you (or any of your subsidiaries, corporate affiliates or agents) initiate
directly or indirectly, or take a direct financial interest in, any Patent
Assertion: (i) against Facebook or any of its subsidiaries or corporate
affiliates, (ii) against any party if such Patent Assertion arises in whole or
in part from any software, technology, product or service of Facebook or any of
its subsidiaries or corporate affiliates, or (iii) against any party relating
to the Software. Notwithstanding the foregoing, 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.
>Facebook or any of its subsidiaries or corporate affiliates files a lawsuit
And then
>against that party
IANAL, but does that mean that if Facebook proper transfers a patent to a new subsidiary (wholly owned patent troll) they can sue you, and the only patent counterclaim you can make without terminating your ready license is against the troll subsidiary, not Facebook proper?
>>
David Recordon
Added Yesterday 21:10
Hi all, wanted to jump in here to let everyone know that the RocksDB team is adjusting the licensing such that it will be dual-licensed under the Apache 2 and GPL 2 (for MySQL compatibility) licenses. This should happen shortly and well ahead of August 31st. I'll leave the history and philosophy around licensing alone since it's generally a complex discussion to have and I'm not sure that it has actually been fully captured in this thread especially vis a vis Facebook's intent.
Hopefully this morning's guidance to PMCs can be adjusted since I don't think any of us see a bunch of extra engineering effort as a desirable thing across the ASF projects which are already making use of RocksDB
Thanks,
--David
Looks like they're working on amending this issue, could very well be a case of legal getting involved and the regular engineers not realizing the change or simply not paying attention. Alternatively, maybe this is just crisis management and they were just hoping this wouldn't happen.
I rejected using React in a project just for this reason. I'll be perfectly honest: I didn't (and still don't) completely understand the implications, but on it's face it sounds like trouble.
Am I reading this right that Apache's unwillingness to use rocksdb under the custom license pressured Facebook into switching to Apache || GPLv2? That is pretty cool!
The issue came out of FB developing a RocksDB plugin for Apache Cassandra. If the Apache Cassandra can't take the plugin into its main repo (which was VP Legal's decision in that thread), I suspect FB would be (is? was?) incentivized to relicense to make that happen.
If Facebook had not included this Patent grant and had released React under only the BSD license then any user would be in the exact situation which everyone is complaining so loudly about being in IF they decide to bring a patent action against Facebook. Specifically, you would be open to being sued by Facebook for violating a patent which they own.
What this grant says is for one specific circumstance (you haven't brought a patent suit against them) and for one specific limited set of patents (those related to React), Facebook will not sue you. If you like that protection then don't sue them. If you decide that you do want to bring a patent suit against them then you're right back where you were to begin with. Your one small limited protection is removed and Facebook can once again sue you if you violate one of their patents - just like they could before you started using React in the first place.
This business about it being asymmetrical is IMO a distraction. What would it mean for it to be symmetrical? That it would only trigger if you sue them for a patent violation related to React? What does that mean? You don't hold React patents, Facebook does. How would you sue them for a violation of a React patent? It makes no sense.
To put it concisely, “why is having a patent license until you commence litigation worse than not having one at all?” The answer is subtle.
Many corporate lawyers operate on the assumption that all open source licenses that do not mention patents (BSD, MIT etc) implicitly grant a patent license. Clarifying this ambiguity is seen by them as harmful — that’s why approval of CC0 at OSI was abandoned[1], for example. Including an explicit patent grant removes the possibility this could be argued in court and is seen as an escalation of the patent conflict by Facebook.
Given many voices at Apache are being quietly guided by corporate counsel, this seems the most likely underlying explanation for the antipathy that's been rationalised out into the open.
I like having the alternative. But the problem is that 100% of your ecosystem needs to not use React. And that's the majority of the value I've gotten from using React: all the components I get from libraries like antd that save me a ton of prototyping time.
So if I understand correctly, by using React, you agree that if you sue Facebook, you'll need to stop using React? And that goes no matter the reason why you're suing them for?
So say Facebook was infringing on one of your patent, you could still sue them, but you'd have to give up React if you did. Is that correct?
Probably they just want to keep to door open and when some oracle comes around, their lawyers can make up some connections to patents facebook has then.
As the other comments say, RocksDB is going to be dual licensed both GPLv2 and Apache. What's the advantage to doing so? If I choose to consume the library via the Apache license, I'd never have to contribute back code; doesn't this invalidate the copyleft of GPLv2?
The copyright owner gets to decide how to license their stuff.
So my guess, is Apache and GPLv2 together makes sense if author wants to give the user the maximum choice in combining the library with other software. That because depending which lawyers you ask Apache and GPL are not always compatible.
So say your code is Apache and you want to use RocksDB, you can pick to use the Apache license for it. Your code is GPL, then you can pick GPL.
> Hi all, wanted to jump in here to let everyone know that the RocksDB team is adjusting the licensing such that it will be dual-licensed under the Apache 2 and GPL 2 (for MySQL compatibility) licenses.
This keeps coming up as a concern with back and forth "it's a big deal"/"it's not a big deal" - so if FB has no ill-intent from this, are there any simple, obvious changes they could/should make to the React license?
Sure, Facebook could do what they did for RocksDB and use the Apache 2 license (and optionally, the GPLv2). People have been asking/telling Facebook to do just that for at least a year.
The reason Facebook doesn't, I assume—and this is just my personal opinion—is because Facebook wants to do evil shit in the future that the Apache 2 wouldn't allow them to do. Otherwise, I just don't see any value for Facebook in all the negative publicity they've gotten over this idiotic patent grant crap they're trying to pull.
And now, for a tinfoil hat moment: even though companies like Microsoft and Google are using Facebook's libraries, it is possible that they have some kind of private deal in regards to the patents clause.
> As noted earlier, the availability of a GPLv2 license does not preclude inclusion by an ASF project. If patent rights are not conferred by the 3-clause BSD and required by ASLv2 then would not these licenses be incompatible?
---
> he has discussed the matter with FB's counsel and the word is that the FB license is intentionally incompatible. It is hard to make the argument that it is compatible after hearing that.
Is it any patent, or only patents pertaining to the "Software" as defined in the license (react)?
I cannot sue Facebook for patents in react or lose my react license, but I could some other patent I own e.g. fizzbuzz that Facebook is violating. Is this correct or is it any patent?
If it is any patent, I cannot believe that was the intent even if that's how Apache Foundation is interpreting it.
If your landlord does something that they would normally be within their rights to do in retaliation to you enforcing some other provision in your agreement, then it is illegal. I'd bet that this was not the kind of statute that they could override in a lease agreement.
I wonder if such a clause is actually enforceable. Are there any actual cases where this clause was invoked?
Thanks to David Recordon this has now been fixed and RocksDB is now dual licensed under Apache or GPL 2. The ball has been started to have the same change occur at React, which AFAIK is still under the old BSD+Patents license. Please click through the OP's link to get the current details.
What will Facebook have to do now - they either have to opensource their infringing software or re-write them convincingly (Wasn't there an Oracle vs Google case about duplicate implementations of a function?)
People should be aware that Atom (the editor) uses React internally, so it's possible you face similar legal exposure without even shipping anything just because you agreed to the terms by installing an editor.
Reading a bit more the thread, it's quite surprising. The assignee is Chris Mattmann. From his webpage, he is not a legal counsel. The only evidence of a problem they show is that BSD alone implies a patent grant but coupled with an explicit patent grant, this is not the case anymore. The other evidence is brought by Roy Fielding who does not appear to be a legal counsel either about a discussion (oral?) with Facebook's legal counsel that the license is incompatible with ASLv2.
The whole decision seems to have been taken by "not-a-lawyer" people with their own interpretations. Doesn't the Apache Foundation have some software lawyers they can ask?
Actually that's an interesting point. Patents must be obtained in every country where you want to enforce them. So if you have a product that will never be sold or used in a country where React is patented, you might be ok. However, the license may still apply and you would need to determine whether you are agreeing to respect the patent regardless of it's validity in your country.
So many of these questions would be easier to answer if anyone knew the patents FB actually has obtained on this technology.
You usually can get patents for "software-related inventions" but there's no huge "software patents" problem in most other places like there is in the US.
I'm perfectly aware of that software patents are not enforceable in country I live in as far as I can tell. At same time I work with companies across the globe so I have to at least think at possible implications of licenses.
I opened the issue, as a member of the Apache Cassandra PMC, because a committer employed at FB has a POC that uses RocksDB as a Cassandra storage engine, and wanted to gauge the viability of upstreaming it.
LEGAL-303 (the issue linked here) was opened to ensure that we were legally clear to do that. The answer of "no" basically means that a project such as Cassandra:
- Can create a pluggable storage engine that enables RocksDB
- Can not have the RocksDB implementation of that plugin in an Apache repo
There were about half a dozen other ASF projects (Kafka, Samza, etc) that had some reference to RocksDB in their source, that likely would have been impacted.
There are about 50 ASF projects that seem to have some reference to React in their source, which (as this HN comment section clearly noticed) is much more heavily used, and seems to scare more people.
The RocksDB team has already re-licensed the project to make it compatible.
Whether or not that happens with React will determine the apache projects' reactions - a product like CouchDB, which uses react for its database user interface, may have to strip it out (or use a drop-in alternative) before their next release, which is probably a LOT of work for someone.
I've never liked react (and not only because of the sketchy licensing, in fact, that's fairly low among my qualms about react) so it's nice to see validation.
I'm sticking with Vue, even if (and that's a big if) it might also infringe facebook patents.
I hate Facebook but I also hate patents so I like this license and wish more projects would use it because lawsuits impede progress, damage the economy, and no matter what the laws are curious smart people will always invent.
I'm sorry but I struggle to take anyone here seriously that thinks that having to not use a specific js framework would be grounds to be legally neutered in a patent litigation case
You're right about Caffe2 [1], but Torch and PyTorch have existed for a long time and don't seem infected [2] and [3]. For completeness, both TensorFlow [4] and MXNET [5] are Apache 2.
Luckily React itself is no longer unique in itself, and project can use the ideas popularized by React with liberally licensed alternatives such as Preact or Inferno.
When I first brought up how misguided people were for embracing React and projects with this license, I was downvoted to hell on HN. But really, everyone, THINK ABOUT IT. This is a company that glorifies and celebrates IP theft from others, and lionizes their employees who successfully clone others’ projects. They’ve built their entire business on the back of open source software that wasn’t ever encumbered with the sort of nonsense they’ve attached to their own projects. And this industry is just going to let them have it, because the stuff they are putting out is shiny and “convenient” and free?
Having known so many people involved with Facebook for so long, I have come up with a phrase to describe the cultural phenomenon I’ve witnessed among them – ladder kicking. Basically, people who get a leg up from others, and then do everything in their power to ensure nobody else manages to get there. No, it’s not “human nature” or “how it works.” Silicon Valley and the tech industry at large weren’t built by these sorts of people, and we need to be more active in preventing this mind-virus from spreading.
By the way, the fact that Facebook is using this on their mostly-derivative nonsense isn’t what should concern you. It’s that Google has decided, as a defensive measure, to copy Facebook’s move. Take a look at the code repo for Fuschia and you’ll see what I mean. Imagine if using Kubernetes meant you could never sue Google?