Open source is a gift economy. Receiving a gift does not form a social contract that entitles you to future gifts. It is not a "rug pull" for someone to stop giving you gifts. The old versions of Redis are yours for all time. No one can take that away from you. In fact, Redis is still giving you gifts to this day, just with a different wrapping. The new license seems perfectly reasonable, given how much companies like Amazon have exploited the gift economy to the point where it threatens these startups survival.
> Open source is a gift economy. Receiving a gift does not form a social contract that entitles you to future gifts
I think op was talking about contributors, who essentially gifted back. One might take offense if they were exchanging gifts with someone, and they open a pop-up store and sell what was gifted.
Also, it's also a tiny bit hypocritical to expect revenue sharing with Amazon without doing the same for contributors.
Cannot speak for the community, but I'd assume that the right thing would be to handover Redis brand to a FOSS foundation.
Anything else just complicates things, like we see today, where Valkey and others are referenced as being Redis compatible (at least as far as RESP is concerned), and we either have specific clients (such as valkey-go), or existing Redis clients, that seem to want to maintain compatibility with most (all?) of them.
Even without a CLA, Redis was originally licensed BSD. Releasing code under a BSD license is making a promise that you are OK with people using your code for commercial purposes. If you as a developer express personal feelings of dissatisfaction that someone is doing with your gift what you gave them permission to do, in such a way that could be construed as reneging on your promise, then that would make you dishonorable and untrustworthy.
> Releasing code under a BSD license is making a promise that you are OK with people using your code for commercial purposes.
That's just a microcosm of the larger issue, isn't it? Contributors : Redis :: Redis : AWS - Redis gave Amazon the permission to host and make money off of Redis, but Redis was evidently salty about the state of affairs. I think contributors have reason to be salty too.
Well in this case, antirez literally promised[1] they wouldn't change the license of the core away from BSD, and then Redis Labs did just that, and now antirez is speaking favorably of that decision.
That post was from literally thousands of days ago and seems to be in relation to some confusion at that time.
Honestly to me you can see the tensions that led to the license change in that post. It’s largely consistent with what antirez has said in the post and in this thread.
Ok so just because something is in the past it's become irrelevant? So no promises are ever worth trusting? The creator of redis LITERALLY said "Redis will remain BSD licensed". And it's no longer BSD licensed.
You can try to be a smartass and add random caveats but that's not how language works.
Imagine if everyone thought like you did: "Sure I promise to do X" (not saying that I mean for the next 5 minutes and will then ignore my past promise)
The post title in its original context is clearly referring to the confusion discussed in the very first sentence: "Today a page about the new Common Clause license in the Redis Labs web site was interpreted as if Redis itself switched license." The title is saying that Redis core's license was not switched to Common Clause at that time in 2018. That's all. It is not titled "I promise that Redis will remain open source forever".
> This is not the case, Redis is, and will remain, BSD licensed.
True, he doesn't explicitly say for how long, but I don't think it is unreasonable to read "will remain" as "will remain indefinitely" and not as "will remain so until we change our minds".
> I don't think it is unreasonable to read "will remain" as "will remain indefinitely"
That's a reasonable interpretation. But it involves an assumption on behalf of the reader, of words that are not there. I think it's a stretch to consider that specific post a "literal promise" by Antirez.
That said, I just did more research and must admit I am completely wrong with regards to the bigger picture there. My genuine apologies. In the HN commentary on that same post [1], the cofounder/CTO of Redis Labs (Yiftach) apparently made a much more direct statement that "Redis remains and always will remain, open source, BSD license". Due to use of the word "always", that I think can unambiguously be called a literal promise that was broken by the Redis company.
People sometimes change their mind about decisions over time. That isn't the same thing as breaking a promise. You can infer a promise from any declarative statement, but that doesn't mean your inference is correct.
I feel like once a software is open sourced the name of that software should also be required to remain open sourced as well, and any closing of the source must come with a requirement of forking the software and changing the name.
It's honestly pretty horrible that a group of people can take centuries of man hours away from the community and tell the community to kick rocks when the software they're stealing would NEVER NEVER NEVER have had the traction it has if it were not OPEN SOURCE SOFTWARE TO BEGIN WITH.
There is a rash of thefts going on in broad daylight, and I think people need to start litigating. Its obscene and soulless and doesn't belong as an acceptable thing to do in society.
It is absurd to call this license change stealing when the previous work is still available under the original license. This is more like someone who is giving to the community, still continuing to give, but with a slightly more strict license. Do you expect someone who is doing philantropic work, gets contributions from others, but later becomes less philantropic to change their name?
> NEVER NEVER NEVER have had the traction
There is plenty of even closed source software which has traction in many domains, let alone software released under a license which as anti-rez points out allows the users to freely run their websites on redis, modify and redistribute code etc.(with the exception of running hosting services like Amazon).
For instance, it would be amazing and a great improvement if there was a top-quality CAD program with a similar license to Redis.
You clearly feel very strongly about this topic; out of curiosity, are you a Redis contributor? If so, was your contribution done via your day job (paid time) or was it an outside/independent contribution?
> I feel like once a software is open sourced the name of that software should also be required to remain open sourced as well, and any closing of the source must come with a requirement of forking the software and changing the name.
Who should enforce such a requirement, and under what mechanism?
Open source licenses are copyright licenses, whereas product names are trademarks. These are separate legal concepts. Meanwhile "the community" isn't even a legal entity at all. As I understand it, there's nothing to "start litigating" here, and no one to litigate it.
Are you proposing that the OSI should change the OSD such that open source licenses must include mandatory trademark assignment to an independent nonprofit foundation? That's a rather extreme view if so.
> Receiving a gift does not form a social contract that entitles you to future gifts.
Certainly, but if you exchange gifts with someone every christmas, have been doing so for a decade, and they suddenly stop giving you one, or give you an unpolished one where they previously cleaned it to a shine, you are perfectly within your rights to be a bit miffed at that, and withold your own.
People can say the new license is reasonable (and by all reasonable measures it probably is) but that’s not how it feels, and that’s the important part.
Open source is not a gift economy, and is in fact a different, and long established social contract. Never has this misplaced metaphor been used to describe open source, nor do the contributers demand any return that amounts to an "entitlement to future gifts".
If a company's survival is threatened by people using its open-source software, then maybe it shouldn't be open source. Open source software (to me) says I should be able use the software for any purpose, including commercial ones.
That isn't really the case here. It didn't start with a company. The company came along independently, effectively took control of an open source project that was already well established. They found that support alone wasn't enough to be profitable on and have been lashing out at the community that built around the original project, including independent opensource clients.
It's not that the company shouldn't have the code as open source. The company shouldn't exist in the first place and we're all suffering for it.
I agree. It seems that "open core"[1] is a failed experiment. I'm hopeful the idea will be replaced either by governments being smarter about the computers, or companies agreeing to sponsor foundations, or... something. But clearly taking venture investment to build OSS and turn it into a profitable enterprise is an idea that has been thoroughly shown to be flawed.
[1] aptly named if you think about the nuclear power analogy :)
It’s the entire premise of public goods, and the sustainable funding model for that would be through government funding or funding of open source foundations like the CNCF. For example, the IETF standards body is founded by the global charitable organization the Internet Society. How do we continue to build on this public goods funding model for projects like this?
> seems that "open core"[1] is a failed experiment.
I don't think the jury is out. If you look at nginx, you won't see a lot of complains over nginx plus. The nginx brand still has a lot of goodwill and a few open source projects seem to be peacefully building upon it without any issue.
> It is not a "rug pull" for someone to stop giving you gifts.
I'm not sure that the developers contributing code from outside the company see it that way. An easy way to tell if contributors see a license change as a rug pull is if the forks that happen after the rug pull get traction.
> The new license seems perfectly reasonable, given how much companies like Amazon have exploited the gift economy to the point where it threatens these startups survival.
Most of the time, though, open source limits growth - and doesn't threaten the original author's startup. This is a really important distinction because there is a trade-off in choosing open source as a marketing strategy - you do potentially face competition from other contributors, users and parasites that will moderate growth.
It absolutely is though. Do you not see any value in bug fixes? Do you think software never needs to be updated? We all know security updates are crucial, and pretending you can just run the old version forever is insane. If I contributed to redis in the past I did so under the assumption that I was contributing to a product that I could actually keep using. But instead they took all those contributions and gave the contributors a huge middle finger by denying them security updates and fixes.
And sure, you can say "just do the bug fixes yourself". But if I knew that was going to be the case, I'd never have bothered to contribute.
> If I contributed to redis in the past I did so under the assumption that I was contributing to a product that I could actually keep using
It sounds like you also have an assumption that the maintainers will spend the effort to maintain your contribution forever, after already spending the initial effort to review and integrate your contribution. This all takes time and money, which has to come from somewhere.
You're asking "Do you not see any value in bug fixes? Do you think software never needs to be updated?" from the contributors' and users' point of view. But the same exact issue exists from the maintainers' point of view.
A lot of people in this thread are commenting as if an open source contribution is a purely altruistic act by the contributor which then solely benefits the maintainers. That's far from the case in reality. Many third party contributors are writing their contributions as part of their day job (meaning they're paid to do it), and the contribution directly benefits themselves and/or their employer by fixing some problem they encountered or adding some feature that they needed.
> It sounds like you also have an assumption that the maintainers will spend the effort to maintain your contribution forever, after already spending the initial effort to review and integrate your contribution. This all takes time and money, which has to come from somewhere.
Correct, that's generally how open source works. You make a contribution, they merge it, and then the assumption is that it is maintained by the maintainers. You seem to act like this is an odd assumption, even though it is and has been the reality for pretty much every open source project ever.
> But the same exact issue exists from the maintainers' point of view.
Right, it's their "job". Maintaining the software. It's pretty much in the name. That's what happens when you accept contributions. They get merged into the software project, and then you (the maintainer) are responsible for maintaining it, or removing the functionality in a subsequent release. That's why code review happens where you consider the maintenance burden, and features sometimes don't get merged.
If this doesn't sound appealing then don't become the maintainer of an open source project.
> Many third party contributors are writing their contributions as part of their day job (meaning they're paid to do it), and the contribution directly benefits themselves and/or their employer by fixing some problem they encountered or adding some feature that they needed.
Sure, and they could've kept it private so it only benefits them and/or their employer. But instead they chose to open source it in the hopes that
1) The maintainers will maintain it
2) It might benefit someone else
That's the whole deal. You're essentially trading your contributions for future maintenance.
I'm well acquainted with open source maintenance; one of my open source projects has been downloaded over 2 million times, and another is imported by over 8000 other repos on GitHub. I'm not "acting like this is an odd assumption" but rather my point was to consider the economic side from the maintainers' point of view, which many people in this thread are completely ignoring.
As you said above, maintainers don't have to accept your contribution, and if they do, they always have the option of removing it in a future release. And if the maintainers use a CLA or other copyright assignment mechanism, they also have the option of changing the license for future releases. So why are you assuming you can keep using future versions of the product forever if you contribute to it and sign a CLA?
> Sure, and they could've kept it private so it only benefits them and/or their employer
But that doesn't actually provide a net "benefit" to the contributor, because then they have to take on the massive burden of maintaining a private fork.
My point in all this is that contributing to open source often benefits the contributor more than the maintainer, and yet people act like it primarily is an altruistic gift from the contributor which solely benefits the maintainer / product owner.
> But that doesn't actually provide a net "benefit" to the contributor, because then they have to take on the massive burden of maintaining a private fork.
Well yeah, that's the point. Trading contributions vs maintenance burden.
> My point in all this is that contributing to open source often benefits the contributor more than the maintainer
I said it provides more benefit to the contributor than the maintainer, but I did not say that it provides no benefit to the maintainer. Your question doesn't logically follow.
That's exactly my point: It still benefits the maintainer. So they benefited from contributors, then fucked them over later, violating the social contract (but not necessarily the software license because CLAs yadda yadda).
Again, my point in response is that the contributor received more benefits (in the form of code review and ongoing maintenance) than the maintainer. We're talking in circles here.
As long as the contributor isn't a cloud provider, they can continue to use the software unimpeded under the new license. Their contribution continues to be maintained by the maintainers. The contributor hasn't been "fucked over" in any way.
If the contributor is a cloud provider, they can choose to enter into a commercial licensing / revenue share agreement and continue to use the software unimpeded. Or they can decline and use/create a fork, in which case any so-called fucking-over is clearly mutual, because the contributor wants to profit from the software explicitly without making its ongoing development financially sustainable for its original maintainers. That's parasitic behavior.
I must ask, have you contributed to Redis? Or are you just expressing outrage on behalf of other people who actually contributed to Redis and may not actually even share in said outrage?
If the latter, why are you so emotionally invested in this topic? And have you ever been on the other side of the coin, maintaining a widely-used open source project?
