Amazon: Not OK – Why we had to change Elastic licensing
elastic.co
elastic.co
No matter how you slice it, I think Amazon is bad for us end-users, and Elastic is good. Elastic could have released ES as closed source, but they didn't, and the OSS ecosystem is better for it. They were hoping to make money off their product, which I don't think anyone can fault them for, but instead Amazon came in and took a bunch of that money while not giving anything back.
Now Elastic is not happy, and I wouldn't be either. As an end user, I'm grateful the circumstances exist that allow companies to make a living from OSS, and I want to encourage that. AWS is the fly in the ointment there, and I don't see how blaming Elastic for not giving us stuff for free any more is anything other than entitled. We should be grateful that ES is OSS at all, and we should want an environment where companies that produce OSS can thrive, instead of blaming them for wanting to get paid for the work that they release freely into the world.
Amazon hinders that, period. I don't think Elastic is in the wrong here, I think Amazon is.
They are moving onto the "Extend and Extinguish" phase with elastic.
Not sure how good they are with eg MySQL and PostgreSQL, anyone know?
This is, perhaps, exploitation but it seems unlikely they'll kill Elastic ever.
Except elasticsearch was created before the company Elastic even existed. They couldn’t have released it as closed source because they weren’t there to release it at all.
It was written by one guy and it was based on previous open source code in Lucene.
I am ok with them making money off their project, but it isn’t like they are owed a billion dollar company for their work.
Because no other companies build on a foundation of open-source software today? I think that describes every company. Yes, the actual core product is the open-source software, not just components of it, but does that really matter?
I don't understand the distinction you're trying to make here.
I don't really know where I fall on this subject. Companies need a route to monetize when developing open source products. It feels like AWS has been closing many ways to do that. Short term it might feel good for us end users, but long term it's probably bad for the ecosystem.
I am happy to let them compete to see who can offer the best value.
That "one guy" is the CEO and founder of the company. You are making it seem like some guy developed an open source database, and a company later came around and built a business out of it, when that wasn't the case.
Most open source to commercial success stories like Kafka, Mongodb, and Elastic do seem to follow similar path.
Any money made from here on out is not based on the work of creating the software, but on helping people use it. If Amazon does a better job of that than Elastic, than they should win the competition.
You're asserting a comparison where none was made.
If that isn't what you meant then perhaps we are in agreement. :-)
I wasn't trying to imply anything about where most of the functionality was coming from.
https://en.m.wikipedia.org/wiki/Elasticsearch
While I still think this phrasing is unfair, it certainly seems to be the consensus. At least when measured by Google search results.
https://www.google.com/search?q=%22Elasticsearch%20is%20a%20...
I slice it this way: as a company that is highly invested in AWS it is easier for us to deploy AWS ElasticSearch service than to use Elastic's cloud offering or set it up ourselves. But that doesn't mean I like it. Or are you talking about a different end user?
You are seriously underestimating the value Amazon provides by "just installing and maintaining" those services. Maintaining a service at the scale they offer is a huge undertaking.
You get the high-availability, the hundreds of engineers working to keep those services up and make them talk to other AWS services easily. You get teams of engineers on-call to react to any failures.
I agree with you that this has a bad effect on the companies that originally created those projects, but I do see a huge value in what Amazon offers.
As someone building distributed systems, I'd think you'd appreciate that merely "installing" Elasticsearch wouldn't simply cut it for the scale AWS operates at.
I wish Elastic would have focused on their differentiated offerings instead and had let go of their iron grip on Elasticsearch itself (perhaps by creating a Foundation around it).
A relevant question is "would they have been that successful if Elastic were 'just another closed source enterprise product'."
Elastic was successful because a lot of companies tried it out for free and then purchased licenses, or because hobbiysts used it on their personal project and then pushed for it at work.
> Elastic was successful because a lot of companies tried it out for free and then purchased licenses, or because hobbiysts used it on their personal project and then pushed for it at work.
That model is not actually incompatible with closed source. You can always distribute binaries with a liberal usage license. And if your model is selling support, that might actually be helpful, since it's even less practical for a 2nd or 3rd party to support software when they don't have access to the source code, so you'd sell more support contracts
I think Amazon's behavior may end up just harming open source, by punishing the idealism that leads companies to try to make commercialized open source business models work.
I'm aware of that, and have even worked with such companies. However, IMHO it's way harder (and less effective) than supporting an open source product. For instance, it's way harder for a 2nd or 3rd party to diagnose and patch a bug if they don't have the source.
The requirement to give something back and/or avoid taking profit from the work of others is something the OSS world has a complicated relationship to. GPL is quite clear that there's a requirement to pass on source changes, if not explicitly to give them back, and many people were outraged by even this limited requirement and instead chose licenses which imposed no requirements at all.
Similarly, people want their work to be used for free by everyone .. but haven't really considered that this results in them working for the Bezos fortune, for free. Or the US military, for free.
There aren't simple clear answers to these questions, only a slowly evolving discussion.
Certainly, and I think that if we want OSS to thrive we need to move towards a future where it's easy for companies to make a return on their investment by releasing OSS. I think Amazon and all the "I provide your software as a service" providers eat into that and hinder that future.
Yes, GPLv3 was meant to fight Tivoization, but it never anticipated providers providing services on an OSS product without contributing significantly, in combination with the companies that develop the OSS hoping to make money off the same hosted service that the former undercuts.
Basically, monetization strategies for OSS are few, and one that is beneficial for both the company and the consumer is providing hosted services. A third company that doesn't have to develop the software is usually a good enough competitor, but since the developer has the obvious support/knowledge advantage, they can still compete. This breaks down when Amazon comes in with its lock-in advantage and sucsk all the money away from the developer.
This is why we're seeing these new licenses, because there's no way currently to be "OSS except Amazon". I think we do need to figure out some way.
Well, the so-called "service provider loophole" was certainly well-known when GPLv3 was drafted. That's why AGPLv2 was created some years prior. IIRC early GPLv3 drafts contained AGPL-style language, but several of the companies involved in the GPLv3 drafting process (such as Google) objected, and those clauses were withdrawn from the final GPLv3.
It isn’t a limited requirements. There is a very real legal risk that using GPL software in an enterprise code base means you have to open source of your entire code base. That is an unacceptable risk for almost any business so GPL software doesn’t get used.
And this, right there, is how MySQL AB was purchased for a billion dollars in 2008: you dual-license your software. Release it under a copyleft license that bigco's don't want to touch (AGPL3 is tempting today!), and offer a commercial license that gives your customers the freedom to use it as they need.
Plenty of companies use Linux... this seems like something people worried about in the 90's but have now generally accepted.
GPL allows you to "use" but not "make derived works".
You're sadly right that many big enterprises purport to believe there's some special extra risk unique to the GPL, but you're wrong to say it's a "very real legal risk"; there's no reasonable basis for believing that at all.
Amazon took no money from them; they competed on potential revenues.
I think people are upset, not because they don't clearly understand Elastic's motivations, but because Elastic is trying to paint Amazon as the bad guy for using the license Elastic offered. Amazon benefited from Elastic's open license, but so did Elastic. Being open source has greatly benefited Elastic's own business and growth.
That isn't to say that Amazon's size and practices around open source aren't cause for concern, just that Elastic come across as very disingenuous when they try to lay all the blame on Amazon while proclaiming how dedicated they are to "openness".
Why should we be mad at Amazon for adhering to the terms of the license that the ES developers chose?
Software isn't born under the terms of Apache 2/MIT/BSD/a similarly permissive license. The people who developed it chose that license.
We shouldn't. But you can't have your cake and eat it too, and say "well these are the terms you chose so why be mad at someone following them" and then ALSO say "hey, you can't change your terms!".
They're their terms, they can change them if they want to.
I haven't said that. And as far as I know, Amazon hasn't either. Have I missed something from them?
You seem to be the only person passing value judgements:
> Amazon hinders that, period. I don't think Elastic is in the wrong here, I think Amazon is.
This is incorrect: Amazon used Elastic per terms of the license. Elastic didn't care for an infringement on their business, so they've relicensed. No one is in the wrong here.
I'm talking about the general sentiment here. Either Amazon have been playing by the rules and Elastic is within their rights to change those rules, so no problem anywhere, or Amazon has been harming a part of the OSS ecosystem and forced Elastic to make an unpopular change.
> Amazon used Elastic per terms of the license
Maybe I shouldn't have used "in the wrong" and said "is the problem" instead. I don't so much care about whether the rules are being followed as I care that more companies are encouraged to release their software as OSS because they can make money for it. That's a win-win situation to me.
But they can't! Tell me how many companies make profit off purely OSS... RedHat maybe? What else?
And even if they can, they shouldn't be surprised when competitors use their OSS for their own benefit because OSS explicitly allows for that. Making money off OSS is a red herring, just because it works in a couple isolated cases, doesn't mean it's a viable business strategy.
Why the scare quotes? Well, I don't mean all profit according to definition, but specifically the "returns for investors" type. Company profits. Technically you can run a sole proprietorship, and make (say) $100k in profit.. or you could structure as a corporation, pay yourself a $100k salary, and make no profit. It's all the same money, but it's two ways of looking at the portion that I would like to describe as fair compensation to a human for the work they do. When I say "at cost", I don't mean that it's fundamentally impossible to make a living working on OSS; I mean it's fundamentally impossible to get filthy rich with it.
And in my opinion that's a good thing. In my experience, "getting filthy rich" / providing outsized returns to investors almost always comes at someone else's expense. Usually the little guy. It happens when the poor sod paying you can't afford to switch to a competitor, so you're able to wring them dry. The counter-argument goes that we need the "filthy rich" incentive to motivate people to make these things. I think it likely increases the rate of innovation, but I think the amount of cool and useful OSS written by people in their spare time is evidence enough that profit is not a requirement in that area, only financial security.
There is a problem, though, where it's currently very difficult to even make a living wage working on OSS (again, or public goods in general). I think can be solved, and I am working on a project trying to solve this (as a volunteer; we could use help). I'll cut it here (I spent far too much time writing this comment already...), but you can read more at https://wiki.snowdrift.coop
Why is it a win for you?
Reducing it to an issue of following license terms is short sighted, it's having negative repercussions on the software ecosystem and it's a dimension that has to be considered beyond merely a discussion on copyleft and the extent of it.
When you choose an OSS license, you're giving permission to any company or person to exploit your product in any way they want, this is how OSS works.
Amazon is not the only one that can do this and I would be surprised if other cloud vendors didn't also offer ES and other popular OSS software to their customers.
Do you expect that just because you created some OSS you deserve some kind of exclusivity on profits made from it?? If you do, you need to understand you need to use a non-OSS license. This seems to be what Elastic has finally realised, but a bit late.
I know that prominent figures in FOSS have expressed the sentiment that you have to suck it up, but you know, the people actually living through this have a say.
Thus, licensing changes and a conversation on their moral standing.
That seems to be the fundamental problem with this whole tempest in a teapot: people have decided on an idea of what "free software" means in their hearts, and many people think it's about "fairness" and "protecting the little guy". That is noble and good, but isn't extensible to an existing large body of software with licenses that clearly spell out how free they are or are not.
But what is great is that if you don't like the state of affairs you don't have to suck it up: you just have to pick a license that is better suited to your goals.
I have a handful of open source projects on my public Github. They fall into two categories for me:
* Software that is trivial, uninteresting, or easy to replicate: these I've released under the terms of the ISC license (2-clause BSD). I have no expectation it will ever come to much, so I'm happy to free it – if it ever turns up in the license file of the iPhone or a Tesla or something I'll say "cool!" (but it won't because it's not that good ;)) Hopefully someone uses it and it makes their life easier.
* Software that is non-trivial, interesting, or difficult to replicate: I've freed it all under the terms of the AGPLv3 and placed a "business use? contact me about the license" note at the top. If I ever decided to work towards building a product around the software (but I won't because it's not that good ;)) I'd look at a dual-licensing strategy, but in the meantime it's out there for anyone to extend and carry forward and build things on. But I know that the AGPLv3 essentially means FAANG will never touch it because the risk is disproportionate for the reward of using it.
This feels right to me. Your calculus may be different so you can license as you'd wish.
It's really not: the license terms are the root of the problem you are pointing out. We can either voice our (righteous, but ultimately pointless) anger or we can try to analyze what's happening and how to fix it. So let's do the latter.
Amazon offers a fully managed ElasticSearch service running on the core ES code because ElasticSearch was, up to this point, released under the Apache 2.0 license which fully supports Amazon's right to do this.
Amazon offers a fully managed MongoDB compatible database called DocumentDB. It is not based on MongoDB – Amazon reimplemented the core functionality but maintained the MongoDB API layer.
MongoDB Inc. makes the forceful point that it is not a drop in replacement[1] but a rather crippled product that lags behind what MongoDB can do and continues to diverge. This is likely very good marketing for MongoDB and probably helps their company succeed :)
Why did Amazon do this? Why would Amazon use the core ES code but go through a more difficult reimplementation for Mongo?
Because MongoDB's core was licensed under the terms of the AGPL3, but all the drivers that implemented the API functionality were implemented under terms of the Apache 2.0 license.
Beginning to see the solution?
And then doing stuff like this:
> When Amazon announced their Open Distro for Elasticsearch fork, they used code that we believe was copied by a third party from our commercial code and provided it as part of the Open Distro project.
That makes it all the more dodgy. The fact that AWS is also the only cloud provider they complain about, and explicitly name others where they don't have these issues with, paints a pretty clear picture imho...
Ten years ago I would be very hesitant adopting ElasticSearch if I knew that they were the only ones allowed to maintain a cloud solution of it. The fact that is was liberally licensed made me less afraid of vendor lock-in.
In my opinion it seems like Elastic wants ElasticSearch to still be perceived as the fully open source project (with all of its good connotations) it once was.
> AWS is the fly in the ointment there, and I don't see how blaming Elastic for not giving us stuff for free any more is anything other than entitled. We should be grateful that ES is OSS at all, and we should want an environment where companies that produce OSS can thrive, instead of blaming them for wanting to get paid for the work that they release freely into the world.
It's okay to release things as non-OSS. It's also okay to release something as OSS first, and then regret later. But it's super weird that they're painting this picture of AWS being a big evil company when they're just doing exactly what is expected. Can't they just say "we're not able to build a company around the liberal license" instead of this "we're such an open company and we love open source and AWS is ruining everything" talk?
That’s my attitude towards most of these license changes or “open core” pivots. These companies want all the good will and community contributions of “open source” while still being able to wield intellectual property protection laws against other companies who dare compete against them on unrelated, commoditized services like hosting.
But surely it's only a conundrum if the primary goal of a software project is for a single company that has the same name as the software project to exclusively make money by selling hosting and/or support for that software while still using an open source license to attract a community of developers to work for you for free. I'd argue that this conundrum is easily resolvable: either have an open source software project for which anyone can sell support and hosting, or have a software company that develops proprietary software and sells it and related services.
Almost every user would have been free to use it exactly as they do today.
I would be much more hesitant choosing an storage solution if I knew the parent company has problems monetizing upon it.
You should be hesitant about choosing any mission critical product where you don't know how the vendor will make money. This is even the case with stable vendors. How many products has Google killed over the years because they could figure out how to make them profitable (enough)?
HN never ceases to amaze me. This is a post on pattern of exploitative and anti-competitive behavior of AWS. Somehow HN crowd found a way to whine about Google. Every single day multiple anti-Google posts on HN front page was not enough.
To me, that's exactly what you're saying you expected from liberal licenses, but it's delivered by a restrictive license, using the restrictions popularized by GPL licenses. This makes ElasticSearch more open source, rather than less, because now anyone who uses it has to "open" their source code. That's the premise of GPLv3 in a nutshell, and I'm hard-pressed to understand how it's a drawback here.
Have I misunderstood and their new license somehow reduces the openness of their source code to the world?
If this was AGPL, I'd agree with you. IANAL, but SSPL is so broad that it could be construed to cover the Linux kernel, which is a no-no. :(
SSPL: you have to share your source code to the Linux kernel if you sell Elasticsearch-as-a-service"
GPL: you have to share your source code to the Linux kernel if you release binaries
SSPL+GPL: you have to share your source code to the Linux kernel if you sell Elasticsearch-as-a-service or if you release binaries
As long as the source is duly published under the merged SSPL+GPL conditions above, anyone trying to bring a GPL case against Elasticsearch-SSPL-on-Linux-GPL may discover that their judge refuses to enforce the GPL's hostility clause, because the SSPL strengthens the intent of the GPL without weakening it in any respect. We'll never know until someone gets to a judge, though, but my armchair speculation bet is that the "can't combine this license" clause won't hold up when the combination increases the total amount of Copyleft in play without decreasing it in any regard. I hope someone takes this to court and doesn't settle, or else we'll never find out :)
(I'm not your lawyer, this isn't legal advice, etc.)
AGPL is okay.
I'm not hiring a lawyer when I'm selecting a few tools to benchmark. I'm not using EC until I've had the license reviewed by a lawyer. Ergo, redis or similar.
By leaving it grey and a little open to intrepretation they know that AWS lawyers will have to advice AWS to avoid it because it might open them up to legal attack.
(And if we're following the "definitions" then ElasticSearch is no longer "open source" since that has a strict definition, but it's probably not so relevant in this discussion.)
I also don't really object to their license choice at all; what I object to is how they're framing the discussion. This license change is all about business: They want to be able to sell their cloud service without competition. That's perfectly okay, but there's no need to hide this. And certainly no need to "shame" Amazon for building a business on top of something Elastic open sourced.
> Have I misunderstood and their new license somehow reduces the openness of their source code to the world?
I think their new license just shows that it's all about business. If they really wanted an open source license which stimulates anyone to share improvements to ElasticSearch they could have picked GPL. As of now, any big company (Facebook, Google, etc) can create an improved internal fork of ElasticSearch which none of the community will ever be able to take advantage of. And why are they fine with Facebook/Google doing this? Because it won't jeopardize Elastic's cloud offering.
In addition, their new license also makes it harder for other people to build businesses on top of ElasticSearch. Imagine that I invest a ton of time and effort into creating a new management layer which is capable of scaling ElasticSearch drastically better. Something completely novel which looks at current trends of traffic and automatically moves shards around. Non-trivial stuff. Well, sorry, there's no way of building a business on top of this idea.
GPL does not require releasing source code or patches except to those whom you give the binaries. If Facebook creates an improved internal fork of GPL code and runs it inside of Facebook, they can keep their sources/patches private and be in compliance with GPL.
There is a practical advantage to upstreaming your patches: because it makes maintenance easier if those patches are accepted and merged upstream, but there's no legal/license requirement to release them for code that you keep internal to your company.
If ES was under AGPL3, AWS' ES server would absolutely fall in scope.
The Facebook example, however, is tricky. You are technically right: the FSF calls this the Service as a Software Substitute (SaaSS) problem[1]. But Facebook can run an AGPL3 project and, as long as they are careful, will not need to expose any changes they make to it.
In practice, Facebook and other big corporations are extremely unlikely to do this. If they get it wrong, they've inadvertently accepted a license they didn't mean to, which could open them up to liability or worse – relicensing things. To prevent this you get a set of approved permissive licenses that are always kosher. If you want to stray off the path there is a set of massive flaming hoops you have to jump through (VP approval? legal review? just the beginning of your fun!) that are designed to convince you that you should chose a different path that involve more permissively licensed software.
[1]: https://www.gnu.org/philosophy/who-does-that-server-really-s...
This. It is too bad they couldn't have satisfactory financial success building on open source and it is their right and perfectly fine to switch to a different model, but their justification as well as the SSPL dual licensing muddle the water unnecessarily.
At least the blogpost clearly states it is no longer open source, but then it goes "it's just definition, we're actually totally free and open, just, you know, not OSI free and open". SSPL software is not free software, it is not FOSS. Calling it "free and open software" is misleading at best.
The legal team for Fedora expressed this well, I think:
> It is the belief of Fedora that the SSPL is intentionally crafted to be aggressively discriminatory towards a specific class of users. Additionally, it seems clear that the intent of the license author is to cause Fear, Uncertainty, and Doubt towards commercial users of software under that license. To consider the SSPL to be "Free" or "Open Source" causes that shadow to be cast across all other licenses in the FOSS ecosystem, even though none of them carry that risk.
https://lists.fedoraproject.org/archives/list/devel@lists.fe...
I think it is this discrimination that makes in not as free or as open as AGPL and GPL.
If SSPL discriminates against a "field of endeavor" that is making proprietary software, or against "persons or groups" that are proprietary software developers, all copyleft licenses do. That was the point. Here's Richard Stallman on the GPL:
> I make my code available for use in free software, and not for use in proprietary software, in order to encourage other people who write software to make it free as well. I figure that since proprietary software developers use copyright to stop us from sharing, we cooperators can use copyright to give other cooperators an advantage of their own: they can use our code.
https://www.gnu.org/philosophy/pragmatic.html
The quote from Fedora perfectly encapsulates the difference between the OSI review process as people perceive it and the OSI review process as it really was. Are we judging license terms, or what we believe to have been in the secret heart of the company that wrote them? What happens when someone else uses the license, as when independent hackers choose GPL, or startups choose AGPL?
Commercial software makers have plenty to fear from GPL, AGPL, and other open source copyleft licenses, because so many of their business models entail slurping up other people's open code, but not sharing their own. Those licenses prohibit building proprietary software with open code, just not if you happen to "compose" a service with network API calls, rather than build a program by copying code snippets, using frameworks, and linking libraries.
Shay keeps claiming "our users" aren't affected, but who's he fooling? They say, AWS cornered them to adopting dual-license SSPL, what's to say they woudln't do an Oracle in the future (like Sun did with Java and continue to do with their DB offerings?). Slippery slope, sure, but it is indeed slippery for a company struggling to compete with competition and seeking predatory avenues as last ditch attempt to stay alive.
I believe, in all my naivety, that Elastic could have created an Elastic Foundation (like Joyent did with NodeJS, who btw didn't throw a hissy-fit at AWS for Lambda) and invited developers from all walks to shoulder the burden of the core software (which they themselves commoditized by F/OSSing it) so that they could focus on SaaS (like AWS).
I'd like to think, Elastic's real problem is they have hard time competing with AWS in terms of pricing for SaaS (of course, AWS owns infrastructure and so it is a tough battle-front), but if they were paying any attention, AWS Elasticsearch Service was very poor in 2015 and continued to remain so for a long time (it sucks less now), but Elastic's own service wasn't up to the mark, either. I think they misplaced their priorities (see GCP's flawless execution with k8s, managed-k8s, and Anthos) and were caught asleep at the wheel when they could have captured SaaS market away from AWS in those interim years (2015-19) by focusing solely on differentiated features and not on the core Elasticsearch software (which was libre and hence undifferentiated).
Of course, Shay and Elastic know better than I do and I am indeed a grumpy developer who's upset, but I want Elastic to give up their misleading messaging viz. 'doesn't affect / nothing changes for our users'. They're being hypocritical and not doing anyone any favours.
> And to be clear, this change most likely has zero effect on you, our users. And no effect on our customers that engage with us either in cloud or on premises.
No, Shay. It does affect the community, who are also the users of the software.
> We created Elasticsearch; we care about it more than anyone else. It is our life’s work. We will wake up every day and do more to move the technology forward and innovate on your behalf.
I see a lot of "We"s and "Our"s. And that's the problem with CLAs and stealing someone else's work. Companies can't tell anymore who's stealing from whom.
On the other hand if you open source something with a license that permits selling the software, well... what do you expect? You gotta hand it to Amazon. They’ve really hustled the industry by hosting open source code. The code is free, literally anybody else could have done this, but Amazon did it especially well.
Amazon benefits from extended tax holiday. Result: lower cost of doing business.
Amazon appropriates FOSS. Result: lower cost of development.
Amazon knocks off successful products, competing with their own partners in their own walled garden. Result: lower cost of product development.
Amazon allows counterfeit products, fake reviews, and other fraud. Result: lower cost of operations.
Amazon uses gig workers. Result: lower cost of labor.
I'm sensing a pattern...
Amazon's success, their prime (pun!) advantage, is built on aggressively avoiding costs normally incurred by other businesses. They perfected WalMart's strategy.
Sure, they've done some clever stuff. Throw enough spaghetti, some of it will stick. Free shipping with Prime membership is akin to Tencent's freemium (genius). And figuring out how to sell excess capacity was cool.
I'm sure a lot of other leaders would share Bezos' tolerance for risk, commitment to long term plans, if only they weren't micromanaged by Wall St.
I certainly didn't grok AWS for way too long.
And there was plenty of concern trolling about free shipping, eg "how long can they sustain this loss leader?!". Very long when you have free capital, certainly more than anyone else. It fortuitously parlayed into amazing customer retention and upselling. What Prof G (Scott Galloway) has coined the rundle (recurring revenue bundle). Proved so effective, in fact, that everyone's now doing subscriptions for everything.
We saw something similar in the 90s where health care prices plummeted as the now winners developed and convinced the public that their monopolies were good. Now they are able increase prices by 15-20%/yr into the foreseeable future.
This is not a prediction (per se), but a statement that we should realize that monopolistic might be good for the consumer in the short term but bad in the long term.
HNers praise how great Amazon is for Graviton.
HNers praise how great Amazon is for leaving Azure and GCP on the dust.
HNers praise how great Amazon is for FOSS project XYZ.
HNers bash Amazon because yet another project made the wrong assumptions how to make money out of MIT/BSD style licenses.
Yep, I am seeing a pattern definitely.
Any time I've seen it, it's glaringly obvious that HN users come down on both sides of the issue.
Pattern-matching detractors of your position as dominant in a particular venue's discussion is a common partisan failure mode. Doesn't mean you have to succumb to it.
On top of that, lots of discussion has become simply low-quality. The comments on technical posts turn into complaining about something not related to the technical content rather about the product. The amount of complaining and whining is through the roof. Mods should look into "Whine Wednesday" type threads to keep the off-topic whinings and complaining invading every single thread.
We should realize that we all succumb to the same biases when communicating in an online forum and exhibit the same tropes - exasperation at the loss of our 'secret hangout', frustration that companies we like get bashed repeatedly, and over-analyzing and drawing broad conclusions from strangers on the internet.
Try to enjoy the good responses and don't get so bothered by the rest! Or maybe there's another community that is more enjoyable out there. Personally I can put up with some of the noise and repeated points like yours because there's still plenty of value for me in these posts. Best of luck.
I do not criticize or defend Amazon's parasitic relationship with FOSS. Frankly, I don't yet see how it can be any other way. I just merely acknowledge the plain truth. And that Amazon is better at this than the other belligerents.
While I'm a very happy Amazon Prime customer, I'd never be an employee or otherwise do business with Amazon. I just feel like there's no way for me to benefit proportionally. Per the parable of the lion's share.
> ...the wrong assumptions how to make money out of MIT/BSD style licenses.
Go on.
I'm hoping someone, anyone will discuss Peter Hintjens' (ZeroMQ) advice.
I have three projects in my back pocket. Once seen, their secret sauce is trivially reproduced. I can think of no way to publish them as anything other than FOSS. Not even as a service. Nor can I figure out how to pay rent working on them.
Which is a pity. These three tools are pretty neat.
Actually outside of the webdev bubble most companies are willing to pay for software too. SolarWinds in the news for all the wrong reasons now but their bread-and-butter was selling things to big corporations that webdevs would demand for free.
If you're referring to how Amazon Web Services was born out of the excess server capacity in their eCommerce service, this has been debunked as a myth [1]
Bezos knew what he was doing.
[1] https://www.networkworld.com/article/2891297/the-myth-about-...
If Linus all of a sudden woke up tomorrow and said "Hey, I just realized that I'm not being paid a cut by literally every single company in Silicon Valley, that is NOT OK, I am going to shift gears and remove non-contributor code and start releasing Linux as closed source from now on", I feel people would be less forgiving than they are to these much less impactful companies. But Linus would be as "right" as they are, arguably more so.
Many of these companies are simply learning that maybe all those "dinosaurs" of the 90s might have been onto something with commercial licensing, which ultimately seems to be what they actually want: to charge money for their software. Sure, it doesn't get you free contributions and ready-made communities, but it gets you money, which is what a company is supposed to do. And that's fine! It's just not Open Source.
> - You may not distibute[sic] this for a fee, not even "handling" costs.
Also I think the GPL was pretty important to the kernel, since many companies, especially in the 90s, probably would have kept contributions private and their code closed without that gentle push.
OTOH maybe eventually most people would have switched to FreeBSD (or whatever free *BSD would have been the "mainstream" choice), just like they switched to Linux in our universe, since they thought that whatever value add provided by some proprietary unix wasn't worth it anymore.
In a hypothetical copyleft-free universe, sure, there would be a lot more companies using OSS to create proprietary products without having to think about what is a derivative work, linking and distribution restrictions. OTOH all those proprietary companies playing the "commodify your complement" game against each other would ensure that the quantity and quantity of OSS would continually be increasing as well, forcing those companies to continually innovate lest they lose their market to the free OSS alternatives. To repeat, hypothetically speaking, as we don't have an alternate universe to run such experiments in.
This should be "quantity and quality"
And in what concerns x86, they would just support it as well, as Solaris did. HP-UX and Aix also have supported a couple of CPU architectures, as did some of the others.
Just Irix was kind of married with MIPS.
Or Windows would just have won the x86 server room instead.
In any case, there are a couple of major surviving GPL projects, all competition against Linux on the IoT space are BSD/MIT FOSS POSIX clones, ironically one of them being sponsored by Linux foundation (Zephyr), so those around in 20 years will get to appreciate how much of GPL will still be left around.
Perhaps you mean when comparing to the “all Silicon Valley companies” I mentioned, but that still leaves AWS, GCP, Azure, Heroku, and every serverless provider out there as direct comparisons to this situation.
Amazon is giving a lot back to the community, though. They are providing a really valuable service when they provide open source software as a service. They aren't giving back to Elastic the company, but it's important to note the difference, because Amazon isn't being a bad actor here. I think it's reasonable for both Amazon and Elastic to act the way they do, and I think the competition between their respective business models will end up in a better set of products available to developers.
In my opinion this is a short-sighted way for Amazon to do business. If Elastic makes every new feature unusable by Amazon, do to licensing restrictions, Amazon's product will fall behind.
That includes nice people like you and me
It includes reprobates like Amazon
The horrid games they were playing with trade marks is part of why Amazon is a reprobate.
I'm really not interested to know who's in the "right" or in the "wrong". I want to know, what's the motivation for opensource if not "reuse my code please"
Then, you can ask "if they get annoyed at Amazon, why open source?" and the answer is "indeed, and now that they realized their mistake they're changing it".
Notably, they're changing it after building a business off the back of many contributors, many of whom expected to be contributing to OSS. Sure, there's a CLA so there's no legal issue, but I'm not sure it's any more morally virtuous than what Amazon's doing. Both are versions of "trying to make billions of dollars off the backs of other people's work".
There's having cake, and then there's eating it. Either you want to retain control over something so you can monetize it to the max, or you want to particpate (and benefit from) the OSS community and build something that benefits everyone.
There are two main differences here.
1. The scope of the change. My understanding is that Elasticsearch may use Lucene under the hood, but extends it in ways and for use cases that Lucene was not designed for. The same can not be said about AWS taking Elasticsearch and running it as a drop-in replacement.
2. Perhaps most importantly, Elasticsearch didn't build on top of Lucene, and then decide to call itself Lucene. If you think there is so little differentiation between the product you built and the product you built off of, that you are better off highjacking the name, then I question if you made any meaningful differences.
3rd BONUS difference: It is my understanding that a large part of the core Lucene team works at (or at one point worked at) Elastic[0].
2. Amazon adds value here by providing hosting solutions for companies using the elastic search software. So it makes sense to call it "Amazon Elasticsearch Service" since that's what it is. I think interpreting this as Amazon built a new competing product but calling it the same name is not the right interpertation. If that's confusing then maybe modifying it to "amazon elasticsearch hosting service" would be the OK thing to do. Not sure if that would make Elastic happy.
3. That's nice of them (really!). Sounds like win-win. But again, it doesn't make anything they do more justifiable.
No one is asking Amazon to share profits with Elastic. Many people do expect Amazon to honor trademarks of other companies. Many people expect Amazon not to package proprietary features as if they were free and open source.
Lucene is a library that makes it easier to provide indexing and searching of "stuff". It's not a commercial product with a sales and consulting team. I can't think of a more apples and oranges comparison.
Elasticsearch is a name of an open source project. Why is calling something "Amazon Elasticsearch Service" a trademark issue? It's not Amazon's fault they called their company after the name of an open source software (the OSS came first btw). Also, IMHO calling it "Amazon Elasticsearch Service" is fair since it represents exactly what it is. Would it better if they instead took the code, made some closed modifications and then released a service around it with a new name? My thought is no.
This isn't unusual, many other open source projects trademarked the name of their project. Google has some guidance on why a project might want to do this.
https://google.github.io/opencasebook/trademarks/#unmanaged-...
Here is a list of open source projects whose name is trademarked by Canonical.
This is a question, and not a informed opinion/suggestion.
I don't think that's clear. I (and the team I work with) use AWS, like so many of us do. (And the ones who don't very likely use Azure or GCP.)
Why do we give money to AWS (and their kin) every month? I'd submit it's because we're getting value from it. If AWS was actually bad for end users then we, as end users, would walk away.
If you were right, and AWS was bad and Elastic is good, this would be an easy problem. But actually, they're both good. The issue is people who paid AWS to host ES instead of Elastic, and you know who those people are? Us. And with reason!
But I did want to add my personal anecdote to serve as a canary-in-the-mine on what effect the status quo could perceivably have on the proliferation of open source software over the long term:
I think we can all agree on the basic premise that having more open source software is a good thing for society.
For the longest time I dreamed of creating my own open source tools and products and simultaneously monetizing it to create a comfortable life for myself and maybe even eventually turning it into something bigger, and leave my mark on the world.
However, as the years past and events like ElasticSearch v Amazon unfolded, I became more and more disillusioned on the realistic prospects of such an outcome.
Today, I'm in the process of building something that would probably see more success in terms of adoption and do more good in the world if it's released as open source software, but at this point I've basically made up my mind to release it as proprietary software to have a realistic shot of monetizing it to achieve financial independence and eventually build a company around it.
Basically I've weighed the tradeoffs and chose to put my own ability to capture the value of what I created over trying to maximize the value my software could create if open sourced.
This was not a easy decision for me to make, but I suspect I'm not alone in having thought about these tradeoffs and reaching these same conclusions. And as more and more people witness the struggles of companies trying to build viable businesses on top of open source software, more and more people could make the same decision, and thus society would be robbed of all the value that having these pieces of software as open source could have created.
I think the chilling effect these kinds of case studies have on the proliferation of new open source software, and the loss incurred by society as a whole as a result, is at the core of what we should be trying to figure out a solution for, not some philosophical discussion around who's in the right or wrong.
No, many SaaS providers have discovered that it’s the back door in the GPL, Google were the leaders in this field. AWS is egregious but they didn’t invent doing this.
They released ES open source so that they could leverage the community + an open source tool (lucene) to build their tool and later on their company.
If not for the open source they would be just a random actor in search and SOLR would likely be a better tool than it is now.
Also by your logic elastic did to lucene what aws did to elastic...
A license is a choice. It means you choose to not gain revenue by directly licensing the IP. Instead, you choose to put the code out there without any further legal obligations on your part as well as those who use that code.
It also means that you have to find alternate ways of making revenue e.g. by providing consultancy, building services or licensing the trademark (which is an entirely different ball game from open sourcing the code!).
The trouble isn't that Amazon decided to use ElasticSearch in their own offering. The trouble is that Amazon simply out-competes ElasticSearch with their own product when it comes to consultancy, services, etc.
To add insult to injury, Amazon made the mistake of leveraging the ElasticSearch brand a few times too many in ways that just rub the ElasticSearch people the wrong way.
Of course, the founders of ES could never predict how successful their product would become after a decade. There are plenty of open source products engineered by commercial companies that never catch the eye of behemoths like Amazon.
I don't know if it out-competes them on those terms exactly, rather than the advantage of "Well I'm already on AWS and they offer an ES service so why not just use that".
If we want to encourage free markets, the GDPR et al. need to be very careful to disincentivize "traffic within different parts of the same entity."
As is, if my regulatory compliance is satisfied through AWS's data and IAM handling, then if an Amazon-hosted service better integrates with those components, it strictly dominates competition.
That's a pretty unregulatable quality, and one easily optimized by Amazon (for itself), and impossibly by everyone else (on Amazon).
This weaponizes data protection regulation into a moat around large everything-and-the-kitchen-sink I/PaaS providers.
There needs to be balance between (a) protecting data & (b) ensuring a competitive ecosystem with multiple viable solutions.
Not only has the cloud vendor already gone through the hoops of getting an enterprise agreement in place. They're also big, and recognised, and know how to deal with Procurement. And Risk. And Compliance. And Legal.
Not saying I like that situation. It does seem unjust that the big guys can just cherry-pick good products and monetise them without giving anything back. It offends my sense of fairness. But that's commercial reality, at least in today's markets.
A real personal example I experienced: at one time, the founding team behind a project I happened to use at work actually told me they didn't want to do the enhancement I was asking for because my particular scale-out needs were too niche and none of their other paying customers need that and they didn't have the engineering bandwidth to build my feature (the opportunity cost for them was too high to abandon building features needed by their other customers).
So I had to solve this scale-out problem by myself - which was painful (we had high opportunity cost too).
In that situation, if my cloud vendor were to say they would solve that problem for me as they would be willing to invest whatever engineering bandwidth required to make it happen, then I would go with them.
Now if that happens a few times, the cloud vendor's service offering will be much superior to the original project founding team's offering.
Over time, the cloud vendor's offering will also be cheaper.
Of course the trick here is to be watchful of being sucked into a lock-in by the cloud vendor. You will have to insist that all of the features they are doing for you are actually open-source and portable to another cloud. Many companies define such requirements as part of their procurement process and audit for it.
As more companies start to push for such guardrail requirements to prevent cloud lock-in, the open-source commercial support model may still have a chance – but unfortunately that doesn't necessarily mean the project founding company will do well.
Aren't cloud vendors generally less willing then smaller SAAS companies to do custom engineering work for their customers?
> The only aspect of this which is "detrimental to society" arguably is the legal requirements
When it's hard to make a living out of open source software, that's detrimental.
Elastic is a feature, event bus is a feature, database is a feature, compute is a feature.
You're kind of right about this, but it's the issue that AWS just has a massive head-start with any client that already uses AWS. They don't really out-compete, they just use their existing vendor lock-in to gain an advantage. And really, by using your dominance in one "market" to gain an advantage elsewhere ends up feeling like a bit of a grey area.
> To add insult to injury, Amazon made the mistake of leveraging the ElasticSearch brand a few times too many in ways that just rub the ElasticSearch people the wrong way.
You're phrasing this in a way like Amazon "leveraging the ElasticSearch brand" isn't a trademark issue. Is "leveraging the trademarks of another company" suddenly okay (as long as you don't do it 'a few times too many') as long as you're Amazon? What if Amazon started selling smart thermostats by "leveraging" the Nest brand?
I'm not sure if it's actually a gray area, since I'm pretty sure leveraging your market dominance in one area to compete in another is illegal anti-competitive behavior. Isn't that what the whole Microsoft antitrust case was about? It's too bad the government pretty much gave up on enforcing antitrust law for 20 years, since it feels like similar practices became normalized due to lack of enforcement.
Leveraging your monopoly to compete in another area is illegal. Leveraging a strong position isn’t and Amazon is a long ways from a monopoly.
That really depends on interpretation, which has shifted over time and continues to shift. IIRC, recent interpretations of some types of anti-competitive behavior have been rather literal and required something very close to a literal monopoly, which has had the effect of neutering antitrust law in all but the most blatant of cases.
My understanding is that it's arguable that it's anti-competitive to leverage market share advantages more broadly (e.g. antitrust law could be used to constrain/break-up a duopoly).
And while there are probably some services on AWS you can’t find at a competitor - that’s mostly a choice (aka profit margin) made by the competitor.
In the EU, it is "dominant market position" (and explicitly so) that's the threshold for the Commission to take competition action, for instance.
(not trying to argue they aren't dominant. just curious)
British Airways only had a 38% market share when they were sued for abusing their dominant market position in 1998 (which was upheld by the Court of Justice in 2007)
Open source works fine, and does quite well as a business model (use it as free advertising).
What has happened here is a plain old case of monopoly.
Once markets are no longer efficient, the model breaks down. Amazon can use its resources to extinguish competition with their own product.
This is why we need antitrust law.
It's not a failure of capitalism, its not a failure of the businesses involved its just what happens when you run a freeish market. That is, things get out of whack to the point we the people feel it is unjust, it would eventually right itself but this would take a long time and likely do more interim damage than its work allowing, so we fiddle with it, hopefully not breaking anything in the process.
If we think of the system as a delicate natural balance that we should try our best not to disturb too much I think we've immediately taken a very specific stance which itself shouldn't be above critical examination. It is, after all, just a social system and all social systems involve some level of design whether we like that fact or not.
In theory, we could conceive of the possibility of an economic system that both preserves the autonomy and independence of its actors while also preventing monopolies from emerging in the first place. Its a hard problem to wrestle with but its preferable to acquiescing to the blind faith in the invisible hand. We should never give up on an effort to understand how we could evolve our current systems into ones that work better (imagine if we took the same stance with technology).
About four years ago we have attempted to get their software . It felt like I was dealing with Cisco sales people circa 1998. They were clueless on how to do a multi hundred thousand dollar deal - think slow, inefficient, inflexible, unwilling to compromise on extra $500 add on that would have ended up being a rounding error.
ugh the pain that comes from negotiating our contract every year. Or the pain that comes from trying to get trial licenses. Or the pain we're seeing now from switching to SaaS.
For the customer, the biggest advantage of AWS' SaaS offerings over a 3rd party's (hosted on AWS) is the billing. AWS Marketplace negates that. Maybe at some cost to the provider, but I just found ScyllaDB and RedisLabs there, so it must be working for some.
No, it doesn't work for RedisLabs. Amazon offers managed Redis called AWS EC Redis and recently someone I knew decided to move their entire Redis (multiple) clusters from RedisLabs to AWS EC. RedisLabs lost hundred of thousand dollars.
While I understand that some people viewed a successful OSS project is akin to Wordpress: lots of hosting providers, rich ecosystems, _and_ Wordpress main company is still making good money out of it; this is not apple to apple comparison (can't compare Redis and Wordpress).
Redis belongs to the group of MongoDB, ElasticSearch, etc.
> You're kind of right about this, but it's the issue that AWS just has a massive head-start with any client that already uses AWS. They don't really out-compete, they just use their existing vendor lock-in to gain an advantage.
Can't client run his own Elasticsearch inside AWS? By installing and maintaining it yourself (or contracting someone to do it for you).
Then I don't see vendor lock-in sense: "We choose AWS to host us, now we have no real choice but to use Amazon Elasticsearch Service". Am I missing something here?
Take for instance Amazon RDS which is a family of managed relational database services. I don't think "Amazon RDS for MySQL" is an unfair use of the "MySQL" trademark, even if Amazon haven't asked Oracle's permission. The reason here is that it's much clearer in the way RDS is branded that it's not endorsed by the database engines it supports, it uses their trademarks to describe the engines they integrate with which seems reasonable in my view. Amazon RDS is still its own independent brand.
"Amazon Elasticsearch Service" crosses the mark in my opinion because it blurs the line between the two brands and in many ways implies that Amazon actually made Elasticsearch themselves.
I kind of disagree here, the main reason it outcompetes is based on the network of linked self serve services in the ecosystem. We spend a ton of money on Amazon in general, and I would not tout thier consultancy as being anything but ok if not underwhelming.
> it outcompetes is based on the network of linked self serve services in the ecosystem
You have to differentiate between business concerns.
First, the concern of those who are willing to buy an IP license because they think the product is useful to them (akin to buying a Windows license key).
ES doesn't make any revenue here. They don't sell IP licenses. That's a direct consequence of putting an open source license on your product. Anyone can just get a copy of the code and spin up their own instance, no strings attached.
Second, the market of those who are looking towards assistance in using the product (consultancy, support, servicing, hosting,...). You can spin up your own instance and do all the work yourself independently. But for organizations, operating software is an expense: often it's cheaper to outsource those costs towards specialists... such as ES offering consultancy services.
The product license and the type of support you want/need are different business concerns. You can choose to an open source product and host everything yourself, you can choose a closed source product and host it yourself, or you can outsource hosting and support to a partner like ES or AWS.
> We spend a ton of money on Amazon in general, and I would not tout thier consultancy as being anything but ok if not underwhelming.
Don't get me wrong here... But the moment you open up your wallet for AWS, you've already contributed to AWS' market position against ES.
Sure, your experience with AWS might be underwhelming, and that's totally valid. But that doesn't matter if you still go ahead and choose to pay for their services.
A market position isn't tied to the quality of service. It's tied to how much potential customers a business can sweep up and convert into hard revenue. The quality of the service is tangential to that.
You can create an absolutely shoddy user experience, and still dominate a market if you happen to position yourself at the right time, with the right product to the right people.
> A license is a choice. It means you choose to not gain revenue by directly licensing the IP. Instead, you choose to put the code out there without any further legal obligations on your part as well as those who use that code.
Open source doesn't mean you can infringe on its copyright and use trademarks everywhere you like.
> It also means that you have to find alternate ways of making revenue e.g. by providing consultancy, building services or licensing the trademark (which is an entirely different ball game from open sourcing the code!).
You didn't read the whole article?: "I took a personal loan to register the Elasticsearch trademark in 2011 believing in this norm in the open source ecosystem."
It does OSS no good if Elastic prevails on merits the day after they go out of business from lost revenue.
The change in ElasticSearch license here is well publicised. If AWS were to continue to incorporate new changes to ElasticSearch it would be obvious they had deliberately violated the terms of the license and it's much easier to pursue a legal case.
In general, anyone can use your trademark as long as they are using it about you or your product. I can say "I like Coca Cola, it tastes great" or even "Coca Cola is disgusting" but if I put Coca Cola on the menu but give you Pepsi, that's infringing.
This is why some projects have generic names and brand names for the commercial version. PhoneGap and Cordova, RedHat and Centos, etc. Amazon can offer a machine and say "This is Centos, it's mostly the same as RedHat", but they can't say "This is RedHat" unless they pay for actual RedHat.
Or in this case, would Amazon not need Elastic's permission to say "we use Elastic" to advertise AWS?
Not if what you're selling is Coca Cola
EKS => Amazon Elastic Kubernetes Service
Kubernetes is a little different here... it seems a bit more nebulous from an installation/instance point of view. It's a bit like saying we use "Linux". Which Linux? Debian? Ubuntu? RHEL? SUSE?
Where they’d get into trouble is if they said they offered a hosted Elasticsearch, but under the hood it was something else. But, even then they could probably say that their offering was Elasticsearch compatible.
The real question is: was AWS misleading customers? I don’t make any claims one way or the other about this specific case. But I wanted to point out that you don’t always need permission to use another’s trademark.
From [1]:
> Nominative use permits the use of a trademark – even in commercial contexts – if it is the most accurate way to refer to a good or service without misleading consumers as to its source.
[1] https://google.github.io/opencasebook/trademarks/#fair-use-d...
But if you're buying a service from AWS the source is not Elastic.
You might be able to say compatible with elastic search. But using the name in your own product name seems unlikely to hold.
I think this is shortsighted on Amazon's part, because it probably wouldn't cost all that much to make a joint offering.
I would be curious to know where those lawsuits went. Because it seems like something that should have been resolved, and for which you could get an injunction.
The problem is clearly that people think they are getting a service supported by ES, when they are getting a look-a-like copy service. Which is what trademarks are intended to resolve.
In hindsight, maybe it would have worked better for ES, had they called the open source product something else, like how centos isn't called RedHat.
Something like this is also asserted in the OP. However, I'm not so sure that is the case. I don't think it's clear at all.
Knowing that Elasticsearch is (was) open-source, I'd assume that I'm getting an AWS hosted installation of Elasticsearch... which is entirely accurate. If you can install the software on your own server, and AWS offers a managed version of it, I have no expectation that the original developers are involved at all.
Ever since the original release, it looks like AWS has been much better at avoiding any mention of Elastic.co. The original announcement Tweet was definitely misleading.
> In hindsight, maybe it would have worked better for ES, had they called the open source product something else, like how centos isn't called RedHat.
I think this is the major problem, and you're right. Elasticsearch was the original trademark and is the accurate mark for the software. They only formed Elastic.co later, and this is where a lot of confusion originates. Elasticsearch != elastic.co
It might be better for AWS to just include a disclaimer like "AWS Elasticsearch Service includes the open source Elasticsearch software, but is not supported by the original developers." Something like that...
Hell yes. I thought it was Amazon's trademark!
I can see both sides of it though. If there are an anti-trust barrier between running AWS and offering major services on top of it, there would be a better overall segregation and likely more innovation overall. On the other hand, putting up a barrier there would be both complex and leaky, and cause missing out on sorts of efficiencies from close integration of cloud platform + services.
We actually were customers of Elastic's offering for a while, but they went down 3 times in a quarter, which was simply unacceptable. We had to switch, and have been okay since. Our bill is also more than half of what it used to be.
The AWS implementation is quite limited in many ways, and there could be a point where we switch back or host it ourselves.
Until reading this news, I never realized ElastiSearch wasn't an Amazon product, I always believed ElastiSearch was Amazon's invention, because of how Amazon employees talk about (always "Amazon ElastiSearch" phrase, often dropping the "service" part of it, so is easy to assume it is "Amazon's ElastiSearch" like "Microsoft Windows")
So it is not just... "a few times too many", if I am interviewing for the company and got extremely confused, how other people wouldn't be confused too? And that is the whole point of trademark laws!
I get that Elasticsearch wants to run their own company, but I really have no sympathy for their arguments here. They released open source software and now are mad that it is taking on a life of its own that they don’t 100% control. That’s the whole point of open source as far as I’m concerned, other people can do stuff you might not have expected with your code.
Now they’re making it more closed going forward, which is fine and is certainly their right to do. But this argument is so bizarre, instead of saying that we tried to do this open source but unfortunately it makes it too difficult for us as a business so we’re closing things off, they’re trying to spin it as they are the true, good defenders of open source fighting against the forces of evil by closing off their licensing further.
Also my initial question was not purely rhetorical, I would assume "Linux" is also trademarked so I'm wondering what is the difference there and why Redhat selling RHEL has not been the same problem.
Also: https://www.infoworld.com/article/2671387/linus-gets-tough-o...
Edit: based on the Linux Foundation link in another comment, it seems they have a clear process for sublicensing the trademark. So I guess Elastic is claiming AWS just launched their ES service without their legal team ever having bothered looking into the trademark? That seems very strange for such a large company.
Red Hat is a top 2-4 contributor to the Linux kernel though, depending on what source (and year) you take a look (e.g. [1]).
The big difference is that Amazon doesn't contribute back. The comparison seems misguided at best.
[1] https://www.phoronix.com/scan.php?page=news_item&px=Linux-Gi...
alias docker=podmanI wonder if Shay regrets that decision in hindsight.
Why the heck would you feel sure of that? What evidence makes you think anything in Amazon's DNA would go in that direction? Amazon is blatantly missing from every open source conversation or ecosystem I've ever encountered. They give essentially NOTHING back except cheap infra as a service, and we pay dearly for it with the losses we get from their predatory behaviour toward any competing products
I use Amazon. But they're garbage for ecosystems.
imho Amazon is only good for Amazon (in the larger timescale), and maybe also for finding whatever economic McGuffin lies at the bottom of whatever race they happen to be in.
The existence and usage of “Open Distro for Elasticsearch” is not debatable. Whether it infringes trademarks is in court and whether a third party infringed elastic’s copyrights is also in court.
But also, that's contributing "out", not contributing "back". It's the difference between working with your neighbours, vs leaving your extra shit around your house on the curb for someone else to take.
The mechanics of how they relate to the groups from which they filter wealth, that matters imho.
When I get frustrated about how they don't really contribute, it's that I'm intensely cynical about the worldview from which their offerings come.
Anyhow, thanks for the generous comment, and the chance to reflect :)
I would hate to be in their shoes, but it brings a valuable lesson to future entrepreneurs: Do fill the "unfair advantage" box in your business canvas.
No, these are services which probably use some kind of search/indexing service in their implementation. They don't provide a database interface.
The closed alternative would be something like Algolia or Azure Search.
In addition to redhat employing a large number of kernel contributors, ElasticSearch is a complete product the Linux kernel is just a piece of the overall redhat product. The kernel in and of itself is useless. Also redhat provides source rpms for every non-proprietary app/utility that makes up the redhat product.
A more comparable situation would be redhat and centos, and to the point that Elastic is making, redhat is very protective of their trademarks with regards to the CentOs project, they have never stood for and would never stand for a situation like this.
Whatever it is it's pretty deep in the weeds. It looks like the intent is for most users to be unaffected; non-AWS cloud providers to be unaffected; even AWS's Elastic Cloud to be unaffected; but AWS has to stop doing something with specific regard to Elastic Search and I can't figure out what it is.
I think the change that you can’t modify the code and use it yourself in production is a big change that is glossed over. ES is now free as in beer. You can look at the code but you can’t touch it or change what it does.
Edit: I was wrong about this. The license itself does not say this, but the blog post seemed to indicate that it was a change. I think it’s an exclusive inclusive or problem.
[0] https://www.mongodb.com/licensing/server-side-public-license
See 2.2 Restrictions: https://github.com/elastic/elasticsearch/blob/master/license...
With the dual-licensing you have a choice of Elastic license or SSPL, but they both have terms you might not like. And if you use the binaries you automatically use the free as in beer Elastic license.
I’m not familiar with this type of license. Any idea how/when this time frame is decided? Is it 3-5 years from software release?
I guess I’m confused by the use of “automatically converts” with a vague timeline. If it’s automatic why isn’t the time of “automatic” conversion more definitively known? What’s the event that triggers the change?
An individual instance of license will say "the covered code is usable under Foo license from Year-Month-Day"
It’s from the day that the code is released under the license and the four years is the max under BSL (so that people know roughly what the “worst case scenario” it a BSL licensed software would be) but can be specified to be shorter by the one releasing code under it.
Edit: https://artifex.com/licensing/commercial/ notably this lets you avoid concerns about integrating GPL with your commercial offering.
Some time ago I changed development model. Public facing version is still Apache 2 licensed. But now there are no unit tests and no integration tests, those are proprietary now. And I extensively use code generator which is also not public.
It is still possible to fork/modify code. Merging pull request is bit more difficult for me (backport stuff to code generator). But it works great and nobody noticed anything.
Practically any serious use of my library has to go through me now. And I am the hero because my code is virtually without bugs. Magic!!! :)
I become disillusioned long time ago. Also people told me several times unit tests do not matter... but in reality they are most valuable part of know how.
To me it seems like a happy medium of being accessible while still protecting your livelihood.
Personally, I wouldn't use a proprietary library/framework unless absolutely necessary. I think it's a great strategy, actually; OP is sacrificing outside contributions while making it much more difficult for someone to just fork the project and bypass them entirely.
I would say this approach fits great for one-men projects.
It's Apache 2. It's as "Open Source" as you can get [1].
I can take a screenshot of Windows 10 and publish it under an Apache 2 license, but that doesn't make the Windows 10 source code "Open Source".
2. Source Code
The program must include source code, and must allow distribution in source code as well as compiled form. Where some form of a product is not distributed with source code, there must be a well-publicized means of obtaining the source code for no more than a reasonable reproduction cost, preferably downloading via the Internet without charge. The source code must be the preferred form in which a programmer would modify the program. Deliberately obfuscated source code is not allowed. Intermediate forms such as the output of a preprocessor or translator are not allowed.
(emphasis mine)
I agree it does not fit strict definition of OS, but it fits Apache 2 license.
Amazon doesn't have any interest in making their version closed because they want the money from hosting. Even if the product isn't that great, it's super easy if I'm already 100% in on AWS anyway (not necessarily reality, but it is an easy conversation to have and the service should be big enough to warrant investment from AWS).
I applaud the stand they are taking and it will be interesting to see how this plays out.
If they're dumb, yes.
As stated in their blog, changes apply pretty much only if you're either embed or redistribute elasticsearch/kibana. And these are two specific use-cases btw.
If you're already a customer, nothing changes.
However there doesn't seem to be many options left now but for Open Distro to become a complete fork of ES.
Also, if you take a closer look at Open Distro, you will quickly come to the conclusion that you really do not want to deploy what drops out of there. The RPM package does CRAZY stuff that made me exhale audibly enough for coworkers to notice - like spawning a postinstall shellscript that `wget`s a .so for/from an optional library that the Open Distro release team put into an S3 bucket, and then `mv`ing that downloaded file (iirc even without any content verification; so the content could be your proxy's captive portal markup, for all they know) into (again, iirc) /usr/lib. That is from WITHIN AN RPM PACKAGE, mind you, where you could and should really just carry that file yourself.
That and other minor troubles with the tooling surrounding the actual product (ES) made me abandon Open Distro fairly quickly. Which is a shame, since a really freely licensed spin of ES with "Enterprise" features would indeed be very nice to have.
The proxy won't connect to bank.com with an invalid cert unless it's configured incorrectly (but the same is true of the OS anyway)
The McAfee-based proxy we have SOMETIMES (I guess it depends on the content-type and the length of the upstream response) renders a kind of "intermediate" HTML document as the response body, where the human user is supposed to click on a link that makes the UA download the originally requested resource from an internal, ad-hoc mirror. I guess that is due to some virus scanning snake oil.
At any rate, what the packages at Amazon did there is just right up in "that is crazy"-territory.
AWS, from what I know, takes security seriously, and given they themselves use OpenDistro internally, this should become a top priority for them.
[0] https://github.com/opendistro-for-elasticsearch/opendistro-b...
Let’s take a quick look at the features that we are including in Open Distro for Elasticsearch. Some of these are currently available in Amazon Elasticsearch Service; others will become available in future updates.
https://aws.amazon.com/blogs/aws/new-open-distro-for-elastic...
https://github.com/opendistro-for-elasticsearch/opendistro-b...
Um, about that...
> When Amazon announced their Open Distro for Elasticsearch fork, they used code that we believe was copied by a third party from our commercial code and provided it as part of the Open Distro project. We believe this further divided our community and drove additional confusion.
More generally I can't understand (and can't stand either) why people keep defending monopolists on HN. Monopolies are bad, morally, economically, in all sort of ways. They fuel abuse and everyone loses in the end but the handful of plutocrats that control Amazon, Google, Facebook, Microsoft etc.
Those are interesting and specific accusations. Got any proof?
The first sentence of that article:
> Back on September 4th, we filed a lawsuit against floragunn GmbH, the makers of Search Guard, a security plugin for Elasticsearch
If we are going off similarities, Napster also continued to distribute their software while their lawsuit involving the RIAA was unresolved. The court still ruled against them eventually.
I am not sure of the outcome.
Amazon's violation of Elastic's trademark is an issue between two companies: Amazon & Elastic. Elastic has the courts available to them to pursue their case.
Elastric's change of license affects the larger open source and technical communities and it's understandable that contributors who supported the open source project are upset when Elastic changes the nature of the relationship.
But Amazon offering hosted Elasticsearch and forking the project is something that I think is okay. It sucks for Elastic but it's good for customers. Amazon is driving the cost of hosted Elasticsearch down closer to its real costs which will always out-compete Elastic who's trying to use their margin to fund development as well. So many businesses fall into the trap of not charging for their actual value and get eaten when someone else is better at their paid complementary services. Elastic's value is the software, not their hosting abilities.
> It sucks for Elastic but it's good for customers.
Everyone is seemingly happy with ES not being able to monetise the product they build for the community, to subsidise the thousands of developer hours spent on it, so their company can save a few dollars. They (you) would rather than money go to Amazon for providing.. nothing to the ES community.
But co-opting open source to grow your user-base and then switching your license because you don't like the reality of what open source actually entails leaves a bad taste in everyone's mouth.
but the elephant in the room is that there are perfectly acceptable ways to make money off of an open source project and HELP them keep afloat by regular DONATIONS/man power/hosting/etc.
Either Elastic's code used by Amazon is indeed stolen proprietary code and no licensing change is needed to obtain reparation, or Amazon is making lawful use of FOSS source code and the question boils down to "if I publish code under a FOSS license, can anybody use it?", to which the answer is obviously yes. And if you'd prefer it to be "no" then don't publish FOSS.
Regarding the trademark, this indeed seems (to my non-lawyer eyes) to be an infringement (or extremely borderline at the very least) but isn't related to licensing.
At the end of the day offering an OSS license becomes less viable when it seems like major players aren’t playing fairly.
It's quite possible that because both companies target developers, architects, etc who're more than able to distinguish between the two companies' offerings... that Elastic's lawsuits didn't go anywhere
Your allowed to sell Apple Macs and advertise them as "Bob's Apple Mac store" without paying any royalties to Apple.
Similarly, Amazon can deploy the open-source ElasticSearch product, and deploy it, unaltered using the trademark.
I don't think you can name your store "Bob Apple Mac Store". You can advertise that you can buy an Apple Mac at "Bob's Computer Store".
This situation is a very fine line between the technicalities that the trademark law was written for. Amazon knows this, and they have the money to set legal precedent in court to have it their way. And so stomping on Elastic's trademark is the route they chose to further benefit from that company's reputation.
If we accept Elastic's interpretation of trademark law, all retail is illegal.
I bought some break cereal at Walmart this morning that clearly displayed a "Kellog" trademark. Walk down any isle of the store, unauthorized use of trademarks as far as the eye can see. NOT OK.
Fair enough. I still don't think Amazon is violating trademark here, but my analogy was unnecessarily dismissive.
I think that this attempt to take a popular open-source project proprietary is going to blow up in their faces. Users will flock to OpenDistro and this will be the beginning of the end of Elastic unless they reverse this decision.
[0] https://finance.yahoo.com/quote/ESTC/ [1] https://s2.q4cdn.com/265747582/files/doc_financials/2021/q2/...
On the Elastic's side they simply seem to be mad that Open Source is working exactly as it's supposed to work in favor of a big company that happens to be a competitor. It's like they want to have discretionary control on who benefits from Open Source and who doesn't.
I believe this is the real driver of this decision. $ESTC's PE is -100 and PC is -2500, they need to drive a lot of business to their hosted cloud or sell many more licenses to support their valuation [^1], and they're not getting the subscriptions they need from the platform add-ons like Machine Learning, APM, and SIEM (43% YoY revenue growth is great, but I don't believe it's sustainable, and this licensing decision suggests neither does Elastic).
Base Elasticsearch and Kibana are sufficient for a large portion of use cases, including mine. Many other of the "ecosystem" tools they sell have other, established commercial or open-source options (e.g. Splunk vs SIEM, Jaeger/OpenTracing vs APM), and these options won't tie you into the Elastic environment.
> I think that this attempt to take a popular open-source project proprietary is going to blow up in their faces. Users will flock to OpenDistro and this will be the beginning of the end of Elastic unless they reverse this decision.
100% agree. We're going to see an open-source fork, whether from Open Distro (which may have too much baggage) or a new, rebranded project, and many users who don't need Elastic's value-adds will flock there.
^1: Admittedly everything tech is severely overpriced right now.
They didn’t invent “elasticsearch” from scratch, rather they took someone else’s codebase (Lucene) and made it better. Fundamentally that’s what AWS did too... they took open source code and improved on it to offer a very popular managed service. Elastic seems annoyed that AWS has executed better on the managed service front but aren’t offering up strong reasons for this being “NOT OK”. Elastic was happy to use code and concepts from others to build their product but seem annoyed when others did the same to them. I don’t get it.
The brand name thing might have more weight but it will come down to if they were truly enforcing the name the whole time they owned it or are just annoyed with AWS. If the name fell into common use they they likely won’t have much luck protecting it.
There are two main differences here.
1. The scope of the change. My understanding is that Elasticsearch may use Lucene under the hood, but extends it in ways and for use cases that Lucene was not designed for. The same can not be said about AWS taking Elasticsearch and running it as a drop-in replacement.
2. Perhaps most importantly, Elasticsearch didn't build on top of Lucene, and then decide to call itself Lucene. If you think there is so little differentiation between the product you built and the product you built off of, that you are better off highjacking the name, then I question if you made any meaningful differences.
As far as I know AWS hasn't contributed any code of note back to ElasticSearch or Lucene.
Imagine putting a sign on your lawn that says people can walk on your lawn and are even allowed to poop on it if they feel like it and then getting mad at them when they do so.
That being said, I support their right to change their license to whatever they like if it helps them survive as a business or for whatever reason they see fit obviously, more power to them.
I think they are saying that when you tell people something is acceptable, then they can only assume it actually is acceptable to you.
I think that is false. Most things that are said assume that the listener will self-moderate. If I have Crohn's or IBS and I post a sign on my front-lawn saying "bathroom free to use for those in need" I'm not expecting you to pull up a tour-bus full of tourists, move into it and use it as housing, or a sex den for turning tricks. I mean, I should clarify my sign, but honestly if you don't meet me half-way with self-moderation, you are the reason we can't have nice things.
Above all, make sure you always leave money on the table, _especially_ if you are the bigger party.
I think this can become a complicated game of accounting though. Did Amazon take more than they need or did they just build a useful cloud service on top of a widespread open and free product that was released intentionally under those terms?
When Elastic chose the Apache license, what was the goal? Was it to allow as many people to benefit from the software as possible? If so, Amazon is clearly advancing that goal, not hindering it.
Or is the idea that Amazon is somehow blocking Elastic from competing in the cloud search space? Elastic is growing quite rapidly and Amazon's use of ES seems to have only accelerated that growth, so I don't really buy that either.
Furthermore consider this: Is Elastic reciprocating the generosity offered by Apache and the Lucene project, to which they basically did the same thing that Amazon did to them?
But you should expecting that, you can't assume the listener will self-moderate.
People are so diverse, I can't assume everybody behave the same, let alone behave the way I want.
I fully support their right to change their licensing, and I understand they may not have thought through the implications of their license -- and I empathize with that. I also empathize with criticism that Amazon isn't doing a great job of supporting the ecosystem that supports them, it would be nice if they did more. And it goes without saying, but I also strongly empathize with the frustration about the borderline trademark infringement that's happening here. That's a completely separate problem.
But I don't like the implication that Open Source licenses are a legal technicality rather than a specific philosophical choice to allow reuse. People don't need to feel guilty about following Open Source licenses, the idea is to encourage reuse -- even by corporations.
We do harm to that movement when we try and backtrack from that philosophy or say, "sure, you have the legal right to reuse the code, but we're going to try and implement social/technical barriers to you doing so." There are plenty of decent source-available licenses projects can use if that's their intent. They carve out exceptions for small-scale reuse while trying to limit companies like Amazon from capitalizing on the ecosystem. And maybe more projects should use those licenses since they more accurately reflect the outcomes that the authors seem to want. There's nothing wrong with having projects that allow only small-scale reuse.
But if someone releases their project as Open Source I'm going to treat it like Open Source, because that's what the movement is about, and trying to reverse the legal progress we've made by constructing new moral barriers in front of reuse is harmful to that movement. When we say that people have a moral right to reuse, adapt, and share our code, we mean it.
But the point of Open Source is that reciprocity of code/money/value is not required. That's literally the scenario that many of us are trying to build.
It feels like the difference here is that you're looking at "someone builds a giant public hosting service off of our code" as an immoral end. But I'm saying that's not immoral, that is an acceptable result.
It's obviously not the result Elastic wanted, and I empathize with that, but... I don't know, maybe we need to educate people more about what Open Source actually means. Maybe we need to encourage more people to use source available licenses if there's a disconnect in how people understand the actual goals of the movement.
We believe that people have an intrinsic, moral right to share and reuse code. Not just good citizens who help build up the system and support us -- everybody.
(I also don't like Elastic as a company, to be clear, and wouldn't shed many tears if they disappeared tomorrow, there's just a hierarchy of dirtbags and they're not near the top.)
As far as encouraging those source-available licenses--that sounds great, except that, in my experience, people with the temerity to offer source-available licenses get treated like shit anyway because they aren't giving away the farm. So I don't know where we go there, either.
I don't know what else I can do as an Open Source developer in my projects and my terminology to imply that when I say, "you can reuse my code for any reason" I actually mean it. I guess traditional Open Source advocates could abandon the entire term and go off and create a brand new movement where we try to make that even more explicit, but people are just going to follow us there and then try to coopt the term again.
> people with the temerity to offer source-available licenses get treated like shit anyway because they aren't giving away the farm
I will call out people who are doing that.
But really, the only comments I have about source available products are:
A) they don't offer all of the advantages of Open Source (although they offer many more benefits than fully closed-source software), and I think that pointing that out is not a moral judgement, just a statement of fact about what the licenses do and do not allow.
B) people who offer source available licenses need to stop saying that they're basically the same as Open Source, or that they're just a subset of Open Source, or that they exist because Open Source has lost its way.
Because the licenses are not the same. All other debates aside, both us at this point in the conversation recognize this, right? You and I are disagreeing about a fundamental philosophy on what rights and moral responsibilities people have around code. You fundamentally disagree with me about whether or not large companies have the right to completely freely reuse permissive code, or whether they have an obligation to pay for it. That disagreement is so large that it affects our attitudes about whether offering large-scale commercial hosting of an Open Source product is moral.
And it's fine that you and I disagree on that point, but we can look at that disagreement and say that clearly your goals when licensing software are different than mine. So to me, it seems pretty reasonable that people who have this fundamental disagreement with the OSI should acknowledge that instead of acting like the Open Source movement is broken. It's not broken, it disagrees with you about the goals are in making code available to other people.
It's not people being stubborn, it's not that the OSI doesn't understand the consequences of Open Source, it's that it does understand the consequences of Open Source and it disagrees with you about what consequences are desirable. The Open Source movement doesn't need shared source advocates to 'save' us, we need them to acknowledge that their goals are different than ours.
Any code that is GPL licensed is Open Source. It might be distasteful to some people to force contributors to sign a CLA, you might get some criticism from some segments of the community, but it's not problematic in a way that means it's fundamentally non-FOSS.
BSL on the other hand is not Open Source, but becomes Open Source at the point where the BSL license expires and is replaced by an Open version.
----
Personally, I might get some pushback on this, but I actually kind of like BSL more than dual licensing. Dual licensing relies on the fact that people find the GPL toxic. It feels much more to me like a temporary solution, and one that only works by kind of dragging the GPL through the mud. Even among people who don't hate the GPL, it encourages them to think of it as a tool to enforce 'fairness', rather than as a complicated way to use copyright to push towards a world where every user has the rights guaranteed in the GPL for every program they run.
TBH, I vaguely suspect that some of the movement towards SSPL is an evolution of people's attitude towards dual licensing, where they thought that the un-attractiveness of the GPL was the point of the GPL, and now feel like it's not living up to it's 'promise'. The fact that Amazon is able to use GPL code to provide commercial services is seen by those people as a bug, not a feature.
Many of the downsides and restrictions around community contributions with BSL are also present in dual licensing because of the implicit CLA requirements in dual licensed projects. So it's not clear to me that BSL is more harmful to community-built software than dual licensing, and given the above trend, it seems a bit more honest (for lack of a better word).
Because dual licensing doesn't really affect companies like Amazon, it kind of encourages people into these arm races where people say that the GPL has failed in its job because some companies don't hate it (again, the point of the GPL is not to be impossible for companies to use). BSL on the other hand is very straightforward, and because it's upfront about its goals, it's not subject to the same kinds of weird arm races and escalations. You release software as proprietary, we all recognize that it's proprietary and that you want compensation for it, and then at some point it becomes Open Source. That's a really simple model to think about and build around.
----
But all that being said, code that is licensed under the GPL is Open Source, period, regardless of what other licenses it is simultaneously offered under.
BSL licensed code before it expires is not Open Source or FOSS: it's proprietary code that later is Open Sourced once a certain amount of commercial value has been extracted from it.
Or is the problem that Elastic can't effectively monopolize that technology which they purposely offered to the world for free? Well, of course not... how can both of those be true at the same time? The choice to release a product as open source is to intentionally prevent it from being monopolized.
"When the service launched, imagine our surprise when the Amazon CTO tweeted that the service was released in collaboration with us. It was not. And over the years, we have heard repeatedly that this confusion persists. NOT OK."
"So imagine our surprise when Amazon launched their service in 2015 based on Elasticsearch and called it Amazon Elasticsearch Service. We consider this to be a pretty obvious trademark violation. NOT OK."
"When Amazon announced their Open Distro for Elasticsearch fork, they used code that we believe was copied by a third party from our commercial code and provided it as part of the Open Distro project. We believe this further divided our community and drove additional confusion. "
This just means their CTO was sloppy, Amazon legal department would have never allowed that tweet.
> "So imagine our surprise when Amazon launched their service in 2015 based on Elasticsearch and called it Amazon Elasticsearch Service. We consider this to be a pretty obvious trademark violation. NOT OK."
This is a trademark violation indeed though IANAL, it doesn't require a change to the license to attack them for that. Definitely an abuse of power by Amazon though, completely not ok as they don't care about paying a fine for that, they have all the money in the world. But again, not related to the license thing.
> "When Amazon announced their Open Distro for Elasticsearch fork, they used code that we believe was copied by a third party from our commercial code and provided it as part of the Open Distro project. We believe this further divided our community and drove additional confusion. "
Elastic was known to mix proprietary and open source code and it got to a point where few people knew what was open source and what was not. Many people were not happy with this situation and elastic.co was abusing the situation to charge paid licenses as people were scared of using proprietary code without knowing. The work amazon did to remove all proprietary code from they fork was actually welcomed by the community though I'm not surprised they missed some as it was really hard to tell.
And Elastic tried going the legal route:
https://searchaws.techtarget.com/news/252471650/AWS-faces-El...
It sounds like their whole issue was about confusion in the marketplace, though, and when someone does an oopsie that results in that kind of confusion, it may not be enough to take care of it quietly, on the side. So it seems now Elastic is making more noise, in an effort to clarify things more publicly.
Surely the legal department would have issued some sort of retraction. Can you find it?
>Definitely an abuse of power by Amazon
Yeah, that's what we're saying.
>people were scared of using proprietary code without knowing...I'm not surprised they missed some as it was really hard to tell.
Amazon is a trillion dollar company that has every capability of doing their due diligence. Sloppy communication, abuse of trademarks and stealing proprietary code are all inexcusable behaviors by a company with the size and power that Amazon has.
You're describing the problem as if it were the excuse. Amazon abused their power, stole proprietary code, abused a trademark, and violated the culture of the open source community whose code they were leveraging for profit. There's no excuse for it, even if it was somehow legal - and I don't suspect it was. I suspect that Amazon knows it's not legal - they just figure they can get away with it.
Sure, but we are not talking about "the intern tweeted something incorrect, gather your pitchforks until they delete it".
We are talking about a prolonged time span where AWS completely abused their massive size and market tower to basically do the legal and PR equivalent of laughing in the face of another company they were using and abusing. Details aside, that is a pretty grim view for the world of software, no?
But the tweet is still up: https://twitter.com/Werner/status/649738362086027265 (archive: https://archive.is/0py42)
Pretty sure legal has reviewed it like a 100 times by now: AWS' taking no prisoners here.
1. Delete it with an apology (admitting guilt is not likely to be great for the lawsuit).
2. Delete it and say nothing (destruction of evidence is not likely to be great for the lawsuit).
I don't understand. If I have an ISP and I offer mysql servers, can't I call that offering "Eznzt MySQL Service"?
Personally, I find that thinking about this issue seems more intuitive when imagining tangible physical products. Imagine that Amazon decides to enter the Cookies as a Service market, and starts launching service offerings with names like 'Oreos by Amazon'. At a glance, would one not assume that this was some sort of collaborative effort between Nabisco and Amazon? I think the average consumer would. And the same probably applies in a situation involving a software product.
Amazon sells Hershey bars through its site. I don’t think it needs to get permission to say “here’s the subscribe and save service to buy Hershey bars.”
I think the confusion is whether ElasticCo is endorsing or part of the service offering. So it should be clear that the offering isn’t by ElasticCo.
Back to the chocolate example, as long as Amazon doesn’t make it seem like Hershey is endorsing their site or offering the product they should be clear. I’ve seen this tucked into the fine print on stuff where it says that just because they are selling Hershey it has nothing to do with Hershey the company.
It seems odd that the company wouldn’t want it to be called AWS ElasticSearch as that’s what it is. ElasticSearch software sold as a service by AWS. Calling it something else is more confusing.
A long time ago I had an open source project to manage mysql replication topologies, and I called it mysql-ha. At some point, they reached out to me about the trademark infringement.
They were nice about it, I did not get a legal notice or anything, just a contact from a MySQL employee pointing me to their policy (as in my response to your example: I could have called it ha-for-mysql), and requesting that I changed the name to make it compliant. I ended up with a full rename (called it highbase) and they were kind enough to give me a one year free subscription to MySQL Enterprise as a token of appreciation for my change.
In way that I think is interesting regarding the AWS and Elastic situation, what MySQL's trademark policy intended was to avoid the situation in which a third party could be confused by a product or project name (mysql-ha in my case) as to believe that MySQL, the company, was behind the offering. So any use of the trademark that made it clear they were not involved (as in the "X for MySQL" vs. "MySQL X") was ok.
Specifically to your example (I think), see "Company, Product or Service Names ", where it states the following:
> Do not use Oracle trademarks or potentially confusing variations as all or part of your company, product or service names. If you wish to note the relationship of your products or services to Oracle products or services, please use an appropriate tag line as detailed above. For example, "XYZ for Oracle database" not "OraXYZ or XYZ Oracle"
I mean, sure. Someone can poop on the lawn.
There is a difference between that, and some business coming along with a dump truck full of shit that they then dump on the lawn, and I'm sure you understand that.
Whats happening is they’re selling the same product; legally they’re entitled to do so.
They’re selling it in a deceptive (perhaps even legally dubious way), and thats not ok; but forget that, this has nothing really to do with being the good guys for open source and amazon being the bad guys, thats just the narrative that the elastic PR folk are putting out.
What’s happening here is being out-competed by people selling the same product, because despite being technically inferior (in my view) the competition can sell more of it more cheaply and not really care about the margins.
So... yes, I’m sympathetic, but this PR dance we go through every time pains me.
Just say it: we’re struggling. We cant compete with Amazon on equal terms, so we’re changing the license to force them to pay us royalties, or stop selling it.
You’re not doing it from the goodness of your heart, and if amazon wasn’t kicking your ass, you wouldn’t care, you’d just be laughing at them “trying to run a cloud version of elastic, ha!”. ...but amazon is very very good at that, actually, and very good at selling it.
Who’s going to judge you for not having amazons scale? No one; but they’re not being dicks, they’re doing their jobs, very successfully.
If you don’t like losing, that’s perfectly ok, no one does... but it doesnt make them bad, it just means they’re better at it than you.
Changing things to preserve your competitive edge is totally ok; but I don’t think its right to spin this us-them AWS is the evil empire narrative; youre in this situation because of the decisions you made, take a bit of humble pie and acknowledge responsibility for it as well.
Why do we need to forget the trademark infringement?
If Amazon is engaging in trademark infringement, lying about their connection\collaboration with the trademark holder, and including commercially licensed technology in an open source fork of a project, they are acting very poorly. Your argument of Amazon just being able to execute better falls flat if these facts are true and it means they're cheating, and that deserves some recognition.
Because that’s what the license they used said.
Is there a difference? The sign never said how much shit could be deposited on your lawn.
[1]https://images.law.com/contrib/content/uploads/documents/403...
[2]https://www.courtlistener.com/recap/gov.uscourts.cand.347725...
And yeah, "Amazon Elasticsearch Service" completely fooled me for about a week until Google searches revealed enough about how elastic.co isn't just a site promoting Elasticsearch, but was a provider of instance configuration.
If elasticsearch had this license from day one, that would be fair enough, but many people do not freely contribute time and effort to improving something which is not freely available to all others (whether individual or large corporation).
Elasticsearch is self-victimising here when they are arguably exploiting FOSS contributors good will (though due to the CLA what they are doing is most definitely legal).
Does AWS ever contribute anything that isn't an AWS integration? I'm not asking rhetorically -- those are the only kind of "contribution" I've ever seen from them.
None of the above appear to be related to AWS specific APIs and offer a small sample of total contributions from AWS employees.
Most free/open source software licences come from a different time. In most cases they are applied because the authors want to do open source and it's expected that the licence is enough to uphold that spirit. But it's not enough and hasn't been for a long time now. The AGPL was created for this reason but oddly developers have gone the opposite direction and "permissive" licences have become the fashion. Many of them are now realising there was a reason for licences like GPL and AGPL after all.
Hence the license change yes?
Seems to me that Amazon has grossly overstepped fair play here.
Apache 2.0, section 6:
> 6. Trademarks. This License does not grant permission to use the trade names, trademarks, service marks, or product names of the Licensor, except as required for reasonable and customary use in describing the origin of the Work and reproducing the content of the NOTICE file.
I feel like I just said this a few days ago: https://news.ycombinator.com/item?id=25796849
The main value of open source to businesses is that support is truly commodified and there is no one with a stranglehold on it. ElasticSearch is trying to remove what makes open source appealing to businesses. No one wants to build their infrastructure on something with expensive IBM/Oracle-costing support. Basically, from now on, ElasticSearch has removed that benefit from their product and businesses are at risk. It's now much less appealing... is the remaining niche profitable? Only time will tell.
Note, why businesses find open-source appealing is not why developers find it appealing, or private individuals.
Redhat didn't become huge because people had all sorts of options for third party support. In fact, I can't say I've ever come across a single enterprise who: uses third party support for their RHEL installed base, has asked for third party support for their RHEL installed base.
No, Thats not true at all. Most open source companies survive off of support contracts. It's why companies choose rhel over centos.
What's stopping you from running Elastic without paying for support under the new license?
Elastic has iterated over and over, taking years to remove obvious problems with their products, building heavily on the community for input about their needs, but still managing to ignore them for a long time. I still remember searching for anything that's not Kibana since their interface has been dreadful (and probably still is). I remember people turning away from Logstash to Fluentd and others pretty early, but don't know the exact reasons. I remember when pretty important and frankly "core" stuff like authentication and authorization among others moved into Shield and other specialized commercial plugins.
They have leveraged almost a decade of developer good-will to cope with their inherent architectural problems and to fight for introducing "weird open source software" in their respective companies and ultimately give them their street cred of "logging aggregation == ELK". Now, after most of their stack "just works" like people expect it to, they throw it all away, putting people who fought for them in license jeopardy while pointing the finger at Amazon? I don't have any sympathy for this. It's your business, if it fails, nobody is at fault but yourself, especially if you a 14B behemoth. May the exodus begin, it's long overdue.
It is not, Elastic does not have a monopoly on what open-source is or mean.
> Our license change is aimed at preventing companies from taking our Elasticsearch and Kibana products and providing them directly as a service without collaborating with us.
Therefore they admit that their market strategy was bad to being with. A lot of open-source technology is being used extensively by AWS/Azure/GCP. I don't see the maintainers of Kubernetes, MariaDB, [insert your favorite OSS project] or even Linux arguing that they are somehow owed money for a product they voluntarily distributed under a permissive license.
> When Amazon announced their Open Distro for Elasticsearch fork, they used code that we believe was copied by a third party from our commercial code and provided it as part of the Open Distro project. We believe this further divided our community and drove additional confusion.
Then sue them and be done with it.
Every time this particular subject comes up I get slightly riled up because it is just a display of misplaced after-the-fact outrage over poor business development planning. Elastic grew because of that open source license and benefitted from a wide adoption because of it. Now that they realized that giving away software isn't a great way to make money they pull this bait-and-switch and expect "the community" to blame the big bad company.
I think trademark violations are pretty bad and a real punch below the belt, but I'm not a lawyer so I don't know if that is actually happening. Amazon also offers Redis as a service, so does Azure. They both have Redis in the name. They also offer MS SQL as a service, however that has a proprietary license which the end customer pays for so it's an unfair comparison. I wonder if the monetisation strategy, which is basically Elastic Cloud, is the best option for Elastic. They are essentially providing a mini managed Elasticsearch cluster which is away from the rest of the infrastructure which development teams are already maintaining. Of course they will be competing with Amazon then and likely going to lose, since Amazon has so much more. Other OSS products have found more lucrative and less costly monetisation models than operating your own cloud hosting provider. I hope Elastic will find a way to sustain themselves in a way which makes the owners happy, because their product is really good.
They do struggle a little bit on their AWS ES offerings if you go across certain threshold.
> I wonder if the monetisation strategy, which is basically Elastic Cloud, is the best option for Elastic.
Redis has RedisLab (cloud) and I can tell you AWS EC Redis does eat some of their customers through various reasons.
After all, US corporate giving, from a cursory search, in 2019 alone was 21.09 billion USD[0] if even 1% of that made it toward open source, that would fund an overwhelming amount of projects overnight. Just 1%. And it would be extremely effective per dollar in terms of what society gets back in return.
I don't know why tech companies don't see it this way in particular.
[0]: https://www.nptrust.org/philanthropic-resources/charitable-g....
But such a system would completely change the dynamics of open source, likely in undesirable ways. Keep the money out, I say.
https://web.archive.org/web/20160610144229/http://www.linuxf...
Now imagine if projects like this could receive real, significant ongoing funding. Maybe charity isn't right, (non profit would be a better word here I think). I see it as a net gain.
It doesn't mean it wouldn't be complicated and add more overhead to certain projects of x size though.
Please do elaborate more though if you think it would strongly result in a net negative.
"Smart companies try to commoditize their products’ complements."
https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/
Not sure what other alternatives are there besides changing licensing.
It is literally the AGPL, with even stronger copyleft provisions. It is anti-proprietary in the strongest conceivable way. How is that not open source? It does not infringe upon, and goes out of its way to protect, the four freedoms.
See in particular items 5, 6 and 9.
There is plenty of vagueness in open licenses. Familiarity over time doesn't make that vagueness go away.
This clearly violates point 9, as it impacts "service" source code, not just the program source code. Also as others have pointed out, what exactly is "service" source code is entirely unclear. But it is clear enough to know it isn't Open Source.
Also, it is clear from Elastic's blog post that they are switching to the SSPL in order to discriminant against fields of endeavor.
You can read more about its rejection https://opensource.org/node/1099 and https://www.zdnet.com/article/mongodb-open-source-server-sid...
1. The OSI definition applies to the license itself, not the company's motivations for its use, so that point is irrelevant.
2. I do not see how point 9 amounts to a restriction, for four reasons:
a. The "other software" is packaged together as a service by the company offering the service, not by the SSPL. The SSPL, in other words, recognizes what already exists. It does not create a new thing.
b. The example given by OSI relates to other unrelated programs simply sharing the same media. The SSPL targets programs that are bound together to provide a service. GPLs already recognize linking, for instance, so how does this not apply as a different kind of linking? It's merely happening at the next layer up.
c. Open source is not a restriction - it is the opposite of a restriction. The entire point of free software licenses is to, as the preamble of the GPLs say, "guarantee your freedom to share and change all versions of a program--to make sure it remains free software for all its users". It beggars belief that integration of a program into a packaged product, modified for and made available over the network, should override that protection. If the SSPL is not "open source" for this reason, then neither is the AGPL.
d. Legal ambiguity is not a reflection on the open-source-ness of the license. All FOSS licenses are "ambiguous" until tested in court multiple times.
It kind of hit home for me because I recently had an issue with an unrelated company that has gotten 100 million+ in funding take advantage of my work by removing my name from the content, openly discredit my work under false claims and attempted to steal money from me multiple times while I've done nothing but help grow their business and ask for nothing in return other than our agreed upon compensation.
What I got from this write up is there's always going to be people and corporations out there who do their best to take advantage of you for the sake of profiting off your work using whatever means necessary, even if it's maybe illegal. I pity companies like this, especially the people who are making the decisions because that's the legacy they are leaving behind and if they happen to have children, they are probably forcing that mindset onto them as well.
(AWS had EC2 before Elastic’s trademark was registered.)
FTA: When the service launched, imagine our surprise when the Amazon CTO tweeted that the service was released in collaboration with us. It was not. And over the years, we have heard repeatedly that this confusion persists. NOT OK.
In this specific case, it doesn't seem tricky to me. When you have concrete examples of people getting confused, no speculation is needed.
As a user, I don't care in the least about the business relationship behind the product. I care about whether Amazon RDS works like SQL Server and whether Amazon Elasticsearch Service works like Elasticsearch. What financial arrangements, if any, are behind the scenes are not a concern to users.
From Amazon's perspective, if they contributed a single fix, or asked a single question of ElasticSearch on the issue tracker, then this is a product born from colloboration.
It's difficult to think anyone is going to think that Amazon ElasticSearch is by anyone other than Amazon.
Doesn't elastic also use Amazon trademarks in their code and documentation? (e.g. ec2, etc..)? I'm not a licence expert, but maybe if you have a have a legal licence to run it, you probably can also name it like that?
Also:
>Our efforts to resolve the problem with Amazon failed, forcing us to file a lawsuit. NOT OK.
This and several other sentences alleging illegal behavior on the part of Amazon seem suspicious to me. When I hear someone say that they had to sue another company, but provide no further details of the suit, then I can only assume that their lawsuit was summarily dismissed by the judge. Otherwise, they'd talk about the litigation - there is no legal condition I could think of where you would be allowed to disclose the existence of a lawsuit and make general allegations about a company, but not disclose the existence of at least a settlement agreement, if not a legal judgment.
Does anyone know if Elastic's Amazon lawsuit went anywhere?
Isn't that just Nominative fair use: referencing a mark to identify the actual goods and services that the trademark holder identifies with the mark?
Especially when it launched and there wasn't a fork.
Sure, hobbyist projects and foundations for FOSS software still exist, but important infrastructure projects like Mongo, Redis, and now Elastic have all recently changed their licenses from "true" FOSS to "some rights reserved".
One might argue that the point of FOSS is not to make money. But GNU/FSF have said repeatedly that it's OK to "sell" your software. How do you do that when a FAANGMO can easily out-scale you and put you out of business?
If I were to start an actual software company, I would give very serious consideration to licenses like Polyform[0] over "true" FOSS, at least for the important, money-making parts where it would be impossible to compete with a FAANGMO.
They are. What they aren't confident in is their ability to differentiate from someone who offers it moderately competently, while not having to pay a single cent in development costs, unlike Elastic, which have to pay for almost all of them.
Could have come across better, but otherwise I support the authors assertions.
Blatantly stealing the trademark, not even entering negotiations. Lying on Twitter about being in a partnership with a company when they are not is the kind of behavior I expect from a shady sneakers reseller on Reddit, not AWS. In my book, this is shockingly unprofessional and indicates some serious rot as a company...
Surprisingly, we found that AWS did better for our use-case. Better IaC - easy to set up clusters with Terraform, and associated alerts. Better monitoring and easier setup. Better price/performance. AWS is obviously lower friction from a purchasing point too once you're already an AWS user.
This makes me curious if Elastic are shooting themselves in the foot a bit here.
Fast forward to 2021 and the AWS version is as seamless as most of their offerings. Works with the VPC, backing stores. You can set it up with CloudFormation/CDK. ES has stagnated.
Given these primitives, do you think one solution to the problem is just what we see here, a new licensing structure for some types of open source? Elastic's move here, attacking the issue through licensing, is one way that this sort of business model is becoming more robust over time and would be instructive for other founders looking to create revenue generating software that is also open source.
As a developer, the main reason I _love_ open source is that I help patch issues or inspect the code to get a better understanding. Which is great because the changes Elastic are making to their license are orthogonal to the value prop for your average developer.
Honestly, I think the biggest issue with Elastic-style licenses moving forward is API compatibility. It is just a question of how much money is at stake for a company like Amazon to go from just operationalizing ElasticSearch to running and maintaining an API-compatible fork, just as they've done with Mongo. It would actually be a bit hilarious if Amazon open-sourced said fork with a more permissive license, given that their buck is usually made off of ops.
It’s too bad that Elastic is no longer open source, but respect the companies choice to close source their stuff.
Will be interesting if Amazon just maintains their fork or abandons it to make something else.
I’m not familiar with elastic as a project and not sure how many community contributions they have, but expect that to shrink as I’m not sure many OSS developers will freely contribute to non-OSS projects.
As for trademark stuff, I expect a renaming like Hudson/Jenkins.
What's the difference to you as a user? Or are you simply concerned about Amazon?
I can’t afford lawyers to determine compatibility today. And I suspect that they would say “not compatible, pay to be safe.”
That’s the difference to me as a user.
Elasticsearch B.V. owes you nothing. The source code is still open source, but you should pay for re-selling or providing hosting services around it. They have salaries to pay. Period.
Too many open source "believers" find themselves out of pocket, taking time away from their families and lives, only for companies like Amazaon and other WAANKs out there to make billions in profit. Time for this to stop. Starve them of your hard work and make them pay if they want to use your software. For sharing knowledge, code can still be freely readable, but should not be free of charge.
see: GPLv3 ban on Tivoization. It makes anything v3 radioactive to any company seeking to make $$$.
I don't mean to make generalities about "open source", but a serious chunk of the community genuinely don't care.
Actually, moves like what elastic is doing are necessary to preserve the FOSS ecosystem.
It’s honestly no different than a leach
That's not what Open Source is.
What's actually happening here is that people disagree with the goals of the FOSS movement, which is fine, but then instead of going out and joining any of the many other movements around software licensing that are better suited for them -- instead of releasing products as source available or shared source or noncommercial-reuse/creative-commons or just under any generally permissive license -- instead they act like this is our problem to solve.
The point of Open Source is not to share knowledge, it's to allow people to reuse/share code. There are other movements that are better equipped to solve your problems if your goal is primarily just to share knowledge. But we're not going to drop everything we've worked to build just to accommodate you.
Nobody is forcing you to be a part of this movement. Nobody is forcing you to release your software under MIT or GPL licenses. You can do whatever the heck you want with the software you build, just leave us alone and stop acting like it's our problem that our movement isn't accommodating your goals.
And we owe them nothing. They have a right to relicense and we have right to complain about it.
If you have an argument then make it but trying to kill discussion is in bad form imho.
I am sympathetic to many of Elastic's complaints, but not this one. If you make an open-core product, you have to expect that others will attempt to make competing, possibly open source, alternatives to your proprietary components.
No, that's just not true. So many users, from small hobby side-projects, to large open source projects, and mega-corps care about the licensing of dependencies, each for their own reason, and will not want to build on top of proprietary software that imposes draconian licensing terms.
It doesn't matter what they say, read the license. It's vague and there is no legal precedent. It's a big risk for anyone who cares about licensing issues for their projects.
If you're using SSPL'd Elastic (or Mongo DB, the risks are the same) for anything serious -- i.e. beyond a hobby, get legal advice ASAP.
SSPL isn't an OSI certified license; many would call it at best a 'shared source' license because of the riders attached.
[DELETED because, as user `gpm` points out, OSI doesn't own 'open source' as a trademark, sorry about that -- the need for legal advice doesn't go away, however.] In fact given their kvetching about Amazon and their trademark, Elastic's cheerleading of open source in this and the original blog post seems to be a bit misleading and doing OSI's trademark a disservice.[/DELETED]
Comically, this is why trademarks exist to prevent people from confusing the market with similar and reused terms.
I think we need a CreativeCommons-like trademark for open source software before it’s too late.
Ironicallly, like many other organizations, Elastic themselves have used OSI's approval as a benchmark for 'open source'[1]:
> Is X-Pack now open source?
> Updated on 2018-04-24 with a link to the Elastic License
> Open source licensing maintains a strict definition from the Open Source Initiative (OSI).
> As of 6.3, the X-Pack code is open under the Elastic License. However, it will not be 'open source' as it will not be covered by an OSI approved license. The interaction model for open X-Pack will be identical to the open source Elastic Stack, including the ability to inspect code, create issues and open pull requests via our existing GitHub repositories.
Something along those lines is trademarked
3. Usage that Require Prior Written Approval 3.1. Distributing software under a license approved by OSI ("OSI Approved License")
Disagree. It's a well standardised term of art: it means what the OSI define it to mean. It's pretty precise, and is not a meaningless marketing term like premium. Similarly free software (and especially Free Software) means what the FSF defines it to mean.
When people start using these terms to mean whatever they feel like it should mean, it muddies the waters for those of us trying to have serious discussions about these topics. Tellingly, such redefinitions are generally broader than the accepted OSI definition, so as to include whatever product someone is trying to push.
I disagree. In fact I'll present the counter example of "myself". I don't agree with the OSIs definition of open source, I think it run contrary to the plain meaning of the term, and is contrary to pre-OSI use of the term. I've argued that numerous times on this forum (and I'm going to avoid repeating these arguments in depth here, just google "gpm Open Source site:news.ycombinator.com"/"gpm Open Source site:lobste.rs" and you will be able to find the arguments).
I have never pushed a product claiming to be open source that does not meet the OSIs definition, nor do I anticipate I ever will, since that seems to be a great way to make discussions about the product devolve into arguments about licensing, which is terrible advertising.
The fact that these arguments usually only come up when there is a reason to argue, i.e. someone has used the term in a way outside of the OSIs definition, does not mean that people only think the right definition of the term is something else when it's for their own benefit.
It's borderline gatekeeping and it irks me to no end.
No, it's a term-of-art. When people muddy the waters and try to undermine the standard terminology of a field, it's not some righteous struggle to liberate a term, it's just an obstacle to clear communication.
In aviation, flap is a precise term-of-art, and is never used interchangeably with aileron, despite that an aileron is plainly a kind of flap (in the colloquial sense). If you adopt your own definition of flap, to refer to both flaps and ailerons, no-one is going to sue you, but no-one is going to know what you're talking about. Your use of the term will be considered not merely different, but wrong.
Similarly, you could try telling a physicist that you consider the words power and force to be interchangeable. They're not going to sue you, but they're also not likely to entertain your deliberate misuse of standard terms.
Are pilots and physicists gatekeeping by being so insistent that you use their terms their way?
The OSI definition is a newer more narrow definition adopted long after the term was in broad use.
And fundamentally, the literal meaning of the words open and source do not have connotations beyond the source being available for viewing.
This seems to indicate that the OSI's precise definition was pretty much there from the beginning, unless they didn't really coin the term open source and it was floating around beforehand. I've heard from others that open source was in use before the OSI definition, so perhaps that's the case.
As I mention in my response though, I don't think the term's early history much matters.
[0] https://en.wikipedia.org/wiki/Open_source#Origins
[1] https://web.archive.org/web/20021001164015/http://www.openso...
I'm not sure the early history of the term really matters. Presumably early aeronautical engineers and early physicists had to all agree on the terms of their field. It's fortunate that they did so, and now that their field is mature, their terms are clear and unambiguous.
Someone without an education in physics might not be able to intuit that physicists use the terms strength, hardness, and toughness in distinct and precise ways.
> The OSI definition is a newer more narrow definition adopted long after the term was in broad use.
I'd say it's a more considered, more precise, more meaningful definition.
> the literal meaning of the words open and source do not have connotations beyond the source being available for viewing
I already gave the example of flap, a precise term-of-art in aviation that reuses a non-technical English word in a way that cannot simply be intuited.
I also don't see that open needs to be a synonym of viewable. I think it's fine that open be used to refer to something broader, as in the Open University for instance. [0]
I tend to capitalise the term, Open Source, to emphasise that I'm using it an a precise way. I do the same with Free Software. Not ironclad, but I figure it probably helps.
With all of that said, I don't think anyone should be permitted to deliberately mislead people when they're pushing a product. It's obviously right that false advertising is forbidden by law.
Personally, any project using “open” in the name that’s not OSI, I pretty much ignore. But it seems to be growing (eg, “open core”, “openai”, stuff like this taking about open with non-open licenses).
It’s getting hard to filter out. One of the benefits I think to CCn is that it clearly lets users know what is and is not allowed. Having OSIn might help with people who don’t read licenses for fun.
That said hosting ES is basically the same as building a carwash, or a gas station, or let's say a printing house. You get the machinery and build your own support services around it.
Even the unit economics are not that different. AWS spent probably millions of dollars to push the marginal price down. The initial cost of procurement for machinery might be zero for ES as opposed to buying a printing press, but none of the aforementioned sectors are limited by the cost of machinery. In case of brick and mortar services the cost of land, labor, construction, and logistics are all a lot more important.
Yes, okay, but what about AWS's advantage, their "moat"? Elastic will never be able to match that. This is the same problem that plagues the browser, phone OS (and other) markets. Google can easily spend a billion USD each year on fiddling with Chrome and Android. Mozilla, Canonical, KDE, and others can't.
AWS has the platform advantage, Google has money.
It seems these market forces virtually force ES to become a "public good" like the Linux kernel. (Or Elastic could try to fork it and stop using any kind of free/open/available license. And try to find business niches.)
But at this point the cat is out of the bag. Likely no amount of license engineering will be sufficient to overcome AWS' advantage.
OSI definition 10: License must not restrict interface, and def. 9. License must not restrict other software it gets distributed with. (So I can't put my encrypted bootloader and verifier into the same thing.)
> the simple requirement that if you provide the product as a service, you must also publicly release any modifications as well as the source code of your management layers under SSPL
This provision is effectively impossible for anyone to comply with in practice. Calling this a "simple requirement" is a barefaced lie.
I also think having people saying “we’re open, but read the fine print” is not good for open source collaboration as it increases confusion and complexity.
Elastic is moving the way of a commercial software company. That’s perfectly fine as it’s their company, but it’s just different than open source.
The idea that a license change made to prevent competition and enable a business model centered around extracting monopoly rents from customers has no effect on customers is ludicrous. It's whole point is to have an adverse effect on customers.
&
grep ( with parallel )
...if it matches your use case, you'll find it trivially outperforms elasticsearch.
Edit to add disclaimer: I work on ClickHouse.
So far the FOSS world seems to be pretending this problem doesn’t exist. Pretending a problem doesn’t exist doesn’t make the problem go away. It makes you go away as you become irrelevant.
There is the AGPL, but it's not quite right. It also has the letters G-P-L in it, which spooks a ton of people still influenced by Microsoft's billion dollars worth of anti-GPL FUD. (I'm convinced you could just rename the GPL and all those problems would go away.)
It has created several.
It hasn't created licenses well-crafted for purposes directly contrary to the purpose of having open source software, because that's not what the open source community is interested in.
> So far the FOSS world seems to be pretending this problem doesn’t exist.
From the point of view of the FOSS world, the issue here is not a problem; creators having an exclusive ability to monetize software as a service isn't a purpose open source is intended to serve; in fact, avoiding the lock-in that results from such exclusivity is a big part of the point.
If the creators get nothing, then why bother? Why slave away to make software just to give free labor to billion dollar companies while you get nothing? Is free labor for Amazon what open source is about?
If open source refuses to adapt to the realities of today's software ecosystem, it will die out... or at least "serious" open source projects will die out and all that will remain is hobbyist level stuff, abandonware, and half-done academic projects.
Personally I do think FOSS in its present form is going to die for most major projects. You'll still see FOSS libraries, building blocks, academic projects, and some major projects that really are large and old enough to have enough real grassroots contributors to keep them going. For major projects in the future you're going to have something more like a shareware model but with source-available.
Nobody creating a new large-scale project today is going to give it a license that they know will result in somebody else productizing it, making a fortune, and giving them nothing. At least Amazon acknowledges where things came from... in some cases the productizers even rename the project and don't even give the author credit.
FOSS and its gift culture ethos just isn't working in today's world. The software market of today is a dark forest.
It absolutely is working the same way it always has (to which “gift culture” matters only around the edges). It doesn't work for people who want to start a business with a business model of using copyright law to extract monopoly rents, but then, it never has, and that's always been the point.
And, yes, it's not, for that reason, a good fit for narrow software entrepreneurship, but that's always been the domain of proprietary software.
What's new is startups building on OSS to build mind share, and then trying to shift to rent extraction while wanting to pretend to still be interested in OSS.
It places really polished products beyond the realm of OSS. If you're fine with that, then there's no problem. Perhaps OSS has achieved its goal, namely creating a free and open software ecosystem for nerds and by nerds.
I can't think of a single OSS project used (directly) by a large number of the general public that does not have a company behind it. I think that says something.
Even if they start out as labors of love, OSS that gets beyond the niche stage tends not to have most work done “for free”, it's done (or paid for) by people/firms who are using the software in their business, but where the software is supporting, not the thing being sold. (Whether the OSS is infrastructure that is invisible to customers, or whether what is being sold is support and professional services tied to the OSS software.)
Yes, very few open source projects ever move out of the fringes of relevance. That's always been true. The idea that there has been some radical change making OSS less relevant is just false; what has happened is that OSS has gotten enough mindshare that people who want to use business models that OSS has never been a good fit want to use OSS as an early marketing gimmick, and then pivot out of it without paying a price for not being OSS. And are upset that people who do care about OSS are calling them on their B.S. when they try it.
I wish all software could be at least source-available and preferably available under even more liberal terms if that could be made to work. That way we could see how things work, learn from things, debug with the benefit of source, port things to different platforms or fix platform problems without waiting for the vendor, contribute if for no other reason than experience, and preserve software after vendors go belly-up without having to resort to emulating old platforms whole cloth.
I also wish there was mainstream adoption of open software for privacy and security reasons. I wish people could use operating systems, web browsers, messengers, and so on whose source could be audited so people could understand privacy implications.
That would all give us more freedom and more transparency, but it also requires a business model to sustain those kinds of projects. As it stands nobody outside geekdom uses open source software because there is no business model to sustain OSS with the degree of polish demanded by end users.
I would love to but the terms within the ElasticSearch codebase on Github are quite confusing. Here's the text of the LICENCE.TXT file.
Source code in this repository is covered by one of three licenses: (i) the
Apache License 2.0 (ii) an Apache License 2.0 compatible license (iii) the
Elastic License. The default license throughout the repository is Apache License
2.0 unless the header specifies another license. Elastic Licensed code is found
only in the x-pack directory.
The build produces two sets of binaries - one set that falls under the Elastic
License and another set that falls under Apache License 2.0. The binaries that
contain `-oss` in the artifact name are licensed under Apache License 2.0 and
these binaries do not package any code from the x-pack directory.
[1] https://github.com/elastic/elasticsearch/blob/master/LICENSE...
Sure, at some point someone started to create a plugin for ES (let's say the security/ACL thing in x-pack, used to be called Shield or something like that), they used the ES API and they used runtime linking. (I have no idea if that's okay or not, has been tested in court or not. I know the US Supreme Court will say something about that in June.) But when developing any feature in that plugin nobody thinks of just that plugin. Folks think about ES as a whole, indexes, shards, documents, terms, maybe even in terms of low-level Lucene primitives.
I think it's practically impossible to wear the OSS and the proprietary hat at the same time. (Or separately but on the same project.)
But if you have a copy of the license and the code and it permitted use of it perpetually, then it can continue to be used. That's my understanding.
Going forward the repository will have a dual-license and the top image on https://www.elastic.co/pricing/faq/licensing can hopefully explain that better.
[Disclaimer: I work for Elastic]
Personally, I don't think so, and I think I should have the right to do so. I wonder how this is different from Amazon behavior here. (I want to make clear that I'm not saying Shay or anybody at elastic is anything. This is for the sake of the example.)
Now foobar's author can stop me from using his project name by registering a trademark on it. But the GPL is working as intented.
At the end, "maintaning" a fork of Elastic is wasted engineering effort and time, it would be better to collaborate. But I personally think Elastic should just ignore Amazon and keep doing what their doing, instead of making their product proprietary.
On the other point: "the GPL is working as intented" yes but not as the authors want, hence the change of license! Nothing wrong with that IMHO.
I do not understand why they're not suing Amazon for a trademark violation. Imagine if they called themselves Elastic Amazon Search... Amazon would send them a cease and desist within a week.
Also, if you don't protect your trademarks, you will lose them! They're playing with fire here by not taking action. [1]
Anyway, send them a cease & desist for misusing your trademarks, Elastic!
[1] https://www.forbes.com/sites/oliverherzfeld/2013/02/28/failu...
The community has invested in Elasticsearch and in Kibana, and we need to keep it open source. really open source.
Check out this initiative: https://logz.io/blog/open-source-elasticsearch-doubling-down...
(disclaimer: I'm a developer advocate at Logz.io) If you also want to take part, email open-es-k@logz.io (or reach out to me @horovits)
Let’s keep Elasticsearch and Kibana open source!
An example, a few years ago I saw a few mac app that show you network metered in status bar and little snitch. I don't know how they did that. I wish I can read their code, even if it's truely OSS.
To me, the value that ElasticSearch give to us is great. And when I do some cool thing, I myself want to share too but I don't want other to take my code and make money off it without contributing back to me.
I think Elastic, as a company, doing a good thing here, and AWS is the bad actor here, they even lie about their collaboration between them and Elastic.
In addition, the fact that API’s are not copyrighted makes this even more in favor of the big enterprises like Amazon, as they can release something with the same API.
Honestly, the problem of how to sustain an open source business in this environment is an open question.
What I’m after is being able to pay a 3rd-party vendor to do all the work of setting up a cluster of machines, deal with HA, backups, upgrades, support etc - but stay out of my data, so that I don’t have to force all our enterprise clients to sign updates to our DPA.
I would gladly pay elastic.co for such a service. The only vendor that I’m aware that does is Cognitect with Datomic Cloud: https://docs.datomic.com/cloud/index.html
That way Amazon couldn’t use their code to provide a SaaS until after 4 years and after that time it would be business as usual and be proper open source.
The license they used now is forever non-opensource, which is a much larger change than what’s merited here I think.
More info at the link below
https://stratechery.com/2019/aws-mongodb-and-the-economic-re...
Now an industry behemoth has decided to directly compete with one of their products. That's tough, especially that it's done in a typically heavy handed way, but... Elastic's reaction seems highly disingenuous, basically a PR dance around "we love open source but we didn't realise it allows competition to eat into our revenue stream".
Elasticsearch is great precisely because it's an open source project, not a product. Otherwise it would be a yet another proprietary black box thingy, with a hefty price tag and a bunch of corporate users. As a project, it thrives, enjoys trust, dedicated community, contributions, enthusiast adoption effect, and so on.
I'm sure they could still make plenty of money from their other products, introduce new ones, maybe even get a huge new stream of support contracts from AWS customers. Instead they decided to cannibalise the main source of their success: their brilliant open source project.
This is like a new modern evolution of internet. Cloud, data, apps, etc all becoming utilities in a highly-automated/converged platform.
For the platform providers, it looks like a new oracle lock-in era, but instead of dotcom boom for the rest, its a net-negative because developer jobs and integrators will find most of their work automated (not necessarily a bad thing, just, whats next?).
--- tangent ---
AI becoming a commodity should really open our eyes. Need to realize that most enterprise and commercial apps facilitate a digital process. Our need to "build", "analyze", or "iterate" through processes are becoming areas of automation / embedded analytics.
Cool potential for new user experiences and capability in new era modern apps. Also wonder what the future need for digital apps even encompasses vs augmented reality experiences.
Would be interesting to compare/contrast, what leads to the difference.
They do the same with lots of products really. Postgres and MySQL too for instance. Also never seen postgres or mysql maintainance teams complain about it.
What are the contextual differences that make it a point of conflict with authors/maintainers in one case but not others?
This is purely my personal opinion.
The other non-open-source-but-wants-open-source-clout is Mongo.
If Redis Labs were to do something stupid with the Redis OSS license, Alibaba and AWS would fork it faster than you can HyperLogLog.
Too help clear up confusion between Redis & Redis Labs, you can read about the various licenses here: https://redislabs.com/legal/licenses/
https://anonymoushash.vmbrasseur.com/2021/01/14/elasticsearc...
At least switching to SSPL might bubble the related trademark issue up to a higher-paid set of lawyers within the Amazon monstrosity, and maybe it'll get resolved.
True genius is always making someone else's blood and sweat into a package that gracefully solves a big pain point and Amazon building up AWS has been nothing but genius.
I'm not sure if I'm reading too much into this but it sort of feels like they don't want/expect to keep offering the proper AWS integration that their elastic.co product has now. I know at work we have something hosted by them in AWS and I assume that's inside our VPC and we'd need that feature to keep using them.
If they do still think that feature is important then saying they "work with other parts of Amazon" feels like it's really under-selling that collaboration/integration with AWS.
The rest of Amazon's behavior aside, there's nothing wrong with cloning a feature. Cloning features is in fact an essential part of competition.
Some other thoughts Where do you draw the line between "cloud", consulting, and just hiring your own engineers to run it. One is fully outsourced and one insourced but the result is the same (someone is managing an Elasticsearch cluster and making money doing it)
It's in Amazon's interest to contribute some of their changes back or they end up with an incompatible product that becomes a less attractive "Elasticsearch" alternative.
The sheer inability of OSI to provide a new-age license that can counter AWS.
Can anybody knowledgeable shed some light on this topic ? Like what OSI license can counter AWS & if there are none why aren't OSI doing anything.
It's no secret AGPL was written to solve "the Google problem". SSPL tried to solve "the AWS problem" with copyleft, rather than just banning the use case, which is what Commons Clause did.
https://drewdevault.com/2021/01/19/Elasticsearch-does-not-be...
Preserving their trademark will forbid Amazon from advertising their service as elasticsearch which may help them find and retain customers.
Elasticsearch should lobby for an antitrust investigation into AWS. Here the market is cloud computing is AWS. This is similar to antitrust in the mainframe market or the PC market etc... However, right now it's not clear what the antitrust remedy will be. In those markets things evolved, most recently from desktop PCs to cloud delivered web apps etc...
Beyond that Elastic needs to innovate or join up with someone bigger.
If so, then my bet is Amazon will begin to treat ES like they do Aurora, namely, they will run their own fork of ES that from this point forward will be a separate code base and will evolve independently but will be "compatible" with anything that would otherwise expect the server to be a normal ES server (like how Aurora is compatible with MYSQL).
If Amazon is infringing a trademark (which indeed seems to be the case in my non-expert eyes), reparation should/could be obtained before a court regardless of the license of the code.
If the author has a problem with his FOSS software being used by an entity he doesn't like then he is in disagreement with the FOSS ideal at its core, this is a perfectly respectable opinion but don't blame it on Amazon.
Does anyone using the above approach have any comments about how well this approach is working for them?
What is the best license to use for this use case? I want to allow as many uses of my code as possible, like on premise deployment, schools, ... But with the exception of using it commercially to run the same service as me.
The SSPL text is still a derivative of the AGPL text, which is copyrighted and licensed under the following terms (from https://www.gnu.org/licenses/agpl-3.0.en.html):
Copyright © 2007 Free Software Foundation, Inc. <https://fsf.org/> Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed.
They do not require in that license to modify the GPL license that you keep the original copyright attribution around.
(IANAL - not legal advice, etc etc)
They're not wrong about Amazon misinformation, use of trademarks, and so on, and should have pursued the legal remedies for this more deeply. Called them out publically, and shamed them like this post attempts to do. But, if it didn't work, tough shit. It has nothing to do with the license. They made a contrat with their community when they choose an open source license.
>We created Elasticsearch; we care about it more than anyone else.
No, you didn't. Elasticsearch is the combined work of thousands of contributors.
Aside: using "Free & Open" in your messaging is a pretty low move, deliberately designed to mislead users.
Which license do you consider an OSS licensed elasticsearch?
If you don't use any of the x-pack features (which used to be paid) you're all good.
AWS forked my project and launched it as its own service
[0] https://twitter.com/tim_nolet/status/1317061818574082050
All for another cool billion.
NOT OK
AWS should build a real serverless alternative to it or buy Algolia or something...
But they usually tend to build their own alternative that integrates better into their eco-system.
The Elasticsearch Service has problems: you can't join the cluster like you can with the normal version, permissions are janky at best, it's slow, and it's expensive. For those reasons the projects my team used it on opted to roll our own elasticsearch cluster which proved a better solution long term beyond the initial annoyances. I say that to say it's not a one to one product, which will probably be their defense.
I would do the same if I were AWS.
If you create an index, you have to pay at least 2,50€ per hour it runs.
All pricing examples were multiple thousands of Euros per month.
Maybe not "better", but useful and with a Apache 2.0 license :)
If Amazon commits to dev work then their project might be the one that survives since it’s actually OSS and more capable of being used in more products.
But if they don’t then it will drift and not be very useful any longer.
Sup with the devil: better use a long spoon.
The list of aphorisms for trading with/under/via Amazon is .. large
open source founders should have some monetisation opportunities. if big companies take everything away it’s not cool
The core legal requirement the GPL puts on distributors is that modifications must also be made open source. That's powerful, and attainable. The core legal requirement the SSPL puts on "third-party distributors" is the requirement that all the source code for that distributor's service must be made open source. First: It doesn't even apply to Elastic. Second: Binaries are discrete; services are networked, often involving many pieces, and there's no strong legal definition for what "service" means in the SSPL.
MongoDB invented it, submitted it to OSI for approval in 2018, then withdrew the application in 2019. Its still not an OSI approved license. Every major linux distro ceased distributing MongoDB upon the relicense, under concerns that its not actually an open source license.
Elastic wants to keep the conversation focused on AWS. Look, I like AWS, but they can be pretty icky, I get that. However, this is not a dichotomy. Elastic betrayed the open source community. They started with open source as a major selling point of elasticsearch, used that selling point to gain traction and users; many of whom did not pay elastic for the service, to be sure. When they had secured a moat of success, they flipped the license to one that is not open source, and now those users are forced to come to elastic for support.
Elastic should be able to make money. In the spirit of that, we just need to be clear: They're effectively no better than, say, Algolia. Yeah, I can read the source code. I can't really change it in a meaningful way. I can self-host (which I can do with many closed-source products). I can't sub-contract a specialist like AWS to manage it for me. Them switching to the SSPL is "fine". They're just not an open source company anymore. This is not an "AWS is evil, Elastic is great" situation; this is a "they're both companies who do some good things and some evil things, but above all else they care about money" situation. There are true open source projects which aren't like this; elasticsearch was one of these, it isn't anymore, and we should focus on supporting products which support their users back, not ones which are built to support The Company.
sums up amazons stand pretty much, they talk about how they are suffering from "Elasticsearches bullying behavior" but in reality Elasticsearch were abusing Elasticsearch in their marketing ("partnering with ES.." lies), they created their own fork instead of offering back to upstream, they partnered up with a company that were stealing enterprise code from ES and selling it as their own product, i mean AWS does not have the privilige to express themselves as opensource evangelists.
https://aws.amazon.com/blogs/opensource/keeping-open-source-...
On a broader note: Elastic Co. used Open Source as a marketing point for their product, but now no longer want to be held to the same standard.
I think instead we are finding that norms and values of traditional open source communities are in some ways contradictory/inconsistent; that there can be competing interests where it isn't true that either one of them is the one that "opensource norms and values" privileges; or that the traditional "norms and values" don't necessary lead to the world that enthusiasts had fantasized about.
In a lot of these discussions, I think the underlying basic thing is that some are alleging, often implicitly, that included in the "norms and values of open source" are that if anyone is making money from value provided by open source, it should be authors of that open source, or at least they should get a cut.
I don't think that is in fact one of the traditional norms and values of open source community. In some ways it's even counter to the tradition.
The actual world/ecosystem around open source has evolved to be very different than the one imagined by traditional norms and values though. Compare to how apache httpd was originally written -- 6 or 8 people, each from a different organization, collaborated on company time each getting paid by their employer, to produce something of value to all of their employers, where the only desired 'profit' was the thing being available for all to use.
That is sort of a stereotypical traditional fantasy of open source. It is of software being created without a profit motive, in an ecosystem where people would contribute to such things on 'company time' (they had a steady salary from some company already). The more people using the software you wrote, the better, and you never wanted a cut of their profits -- that is the fantasy of traditional open source norms and values.
That is not the world we ended up with though.
So the problem is that now it is "obvious" to some people that if we wrote the the thing, and then formed a company around that thing we wrote -- it's not "fair" if someone else is making money from it without giving us a cut.
But this isn't a value encoded into open source licenses at all, and that wasn't an oversight, it was intentional because this wasn't in fact a traditional "norm and value" of open source at all, and in fact it is in some ways counter to the actual traditional norms and values, one of which I would say was: Your desire to make a profit from this code should not in fact be allowed to prevent anyone else from using it. It is ElasticSearch which is acting contradictory to norms and values of open source in believing nobody should be able to use their software without giving them a cut of profits from it.
These disputes will keep happening, not because some companies are violating the "norms and values" of open source, but because the actual traditional norms and values of open source are increasingly unable to power a sustainable economy where people can get paid (in the manner they think they deserve?) while producing open source.
But even if they contributed nothing at all, that’s part of open source in that it’s free for everyone to use regardless of anything else, depending on the license.
I think it’s a virtuous byproduct that all this free, allowed use leads to people contributing more open source. Not because of compulsion, but through a shared philosophy.
[0] https://aws.amazon.com/opensource/ [1] https://news.ycombinator.com/item?id=9358843
Is this considered against typical open source norms and values?
If not, what makes this situation different?
So one difference people talk about is that your stack might be based on postgresql and you might sell a service, but you aren't actually selling postgresql as a service. OK... but I suspect there are people selling (especially) postgresql or mysql as a service, without ever sending patches back; say, a traditional kind of PHP web host, right? This hasn't to my knowledge led to much controversy; or the idea that their entire webhosting/dashboard/management layer has to be open source if they provide postgresql. What's the difference?
I am not saying there can't be a difference, there are all sorts of differences always. But in elucidating what the pertinent/meaningful difference is, we actually are clear about what we think, instead of just a gut-reaction "I don't like amazon and I don't think they should be able to profit off of elasticsearch" -- cause that IMO doesn't have anything to do with "opensource values and norms".
I think, again, is that the real problem is that the traditional models of open source, the traditional norms and vlaues of open source, are becoming less and less capable of supporting sustainability and proper income for open source development. (Which reminds me of the OpenSSL problem of course. Is everyone who uses OpenSSL, which means like everyone, violating "norms and values" if they don't send patches back? Obviously not).
Regardless very expensive.
No seriously. Hear me out.
If you are a proponent of capitalism then this is how the system works.
The little fish grow into big fish. The big fish eat the little fish. The ecosystem suffers.
It has always been this way. Many of us remember Microsoft in the nineties. Fewer will remember the phone or oil industry doing the same.
Don’t fight this issue. Fight the system that tolerates this pattern. Money in politics, high cost of litigation are both the real concerns.
I'm saying large industry has undue influence over the regulators and so action only comes when companies like Elastic have gone bankrupt.
And about the ElasticSearch, they should have just used a different license.
That's a strong claim - evidence?
> And about the ElasticSearch, they should have just used a different license.
Hindsight and all that. Not saying I fully agree with Elastic's approach, especially the license uncertainly it creates as noted elsewhere in the thread. But Amazon seems to have gone beyond "hey, this open source, nothing to stop us offering it" here. From The CTO's suggestion Elastic was a partner, to questionable trademark infringement, to potential copying of closed source code. If you read the article, it's notable that Elasticsearch continue to have working relationships with Azure and Google among others.
So there's more to this than just "should've used a different license".
Humans are flawed, the systems they build will be flawed, flaws will wax and wane with circumstances, possibly waxing to the point of intolerability and then they break and are replaced with something else.
Paradise remains fundamentally always out of reach.
And yet there are good moments, good relationships, little pieces of life that are priceless. So I think it makes sense to carve out little niches in life where things work nicely or to try to make things work better on small issues.