[lukaszwojtow]

All this outcry about license switch coming from "community" feels funny. After all, if there is the "community" then they can take the last open-source version and keep developing it themselves, right? But most "communities" are about "take, take, take", not "work, work, work". They often upset only because someone declared they aren't going to work for free any more.

[jzb]

Author of the article here. There may be some scenarios where there's a company just tossing code over the wall under a FOSS license and people complain when it stops. This scenario is not that.

The company now known as Redis did not invent Redis, it started as a company trying to make money hosting other peoples' work. After it finally hired the creator of Redis, it specifically promised not to do what it has just done (move away from three-clause BSD as the license for Redis core) at least twice.

In the development cycle from 7.0.0 until a few days ago, Redis isn't even the majority contributor to the codebase. The largest single contributor is from Tencent. (All of this is in the article.)

If Redis had been doing all the development, had not promised it wouldn't move away from the license, then I might agree that people have little to complain about.

But this situation isn't as you've suggested here where a community is all about "take, take, take" from a company that's been doing all the work. The company was founded on the idea of trying to do what it now complains about Amazon doing, and their claims that cloud companies do not contribute is clearly false -- just look at the code contributions.

[tayo42]

What did that guy and tencent contribute to so much of recently?

[tayo42]

to answer my own question, i didn't realize tencent had their own cloud offering with all the major software available a service, guess they/him just do general development and bug fixes.

[AnonymousPlanet]

In this case the community is the biggest contributor to Redis. The ones that "take, take, take" is Redis the company. Your comment seems way out of place in this light.

[lukaszwojtow]

Good. So now Redis Inc is in trouble because they have to replace community work with their own. If community does most of the work, then what's the problem?

[oefrha]

The problem is too many people are announcing OSS forks so it’s hard to align development efforts and users are confused. No one’s begging Redis Labs (which didn’t create Redis in the first place and only took over the brand with VC money when it was already popular) or whatever they’re called now to keep the bug fixes rolling. They only account for 20-50% of recent development anyway (50% if you attribute all “unknown” contributors to them), with the other 50% from (predominantly Chinese) cloud companies allegedly “pirating” their software, according to some.

I don’t typically ask people to RTFA because that’s against the rules, but you would have known all of the above if you bothered to read the article.

[endisneigh]

What you’re describing isn’t a problem. Why does it matter if there are too many forks? Development also doesn’t need to be aligned to begin with.

It’s like complaining that there are too many implementations on GitHub of the same thing.

[ufocia]

> It’s like complaining that there are too many implementations on GitHub of the same thing.

You're spot on. People are bellyaching that the world doesn't operate according to their arbitrary rules.

Perhaps I'd be happier in a geocentric universe, but it doesn't make a non-geocentric universe bad per se.

[pjmlp]

Yeah, it is incredible how the whole free software movement turned into a bunch of entitled folks that want to be paid for their work, while refusing to put down any penny for the folks that make their tooling possible in first place.

At the same time big corps use it as carte blanche to basically pirate software in a legal way, while following the letter of the licence.

Going back to the open core/demo versions (aka Shareware/Public Domain/Trials) is the only sustainable way to make a living.

[chii]

> Going back to the open core/demo versions

aka, just sell software, rather than make it open source.

What is being balked at is the idea that you can use open-source as a foot-in-the-door marketing and growth hack, which you then reap after some level of popularity/network effect is reached. Some call it bait and switch.

Blaming big corps for "leeching" is just self-serving. They are doing exactly what the license allows them to do - a license for which was chosen at the start to allow for it! If you expected to be paid to make this software, don't opensource it.

[ufocia]

Or perhaps open source it in exchange for being paid, something that developers working for corpos which contribute to (FL)OSS already do.

[AnonymousPlanet]

None of what you say is happenening in this case. Unless by "entitled folks" you mean Redis Inc.

The community has been doing the heavy lifting over the years and Redis Inc has been trying to reap the benefits off of that by providing the software as a service. Which the community was fine with. Turns out other companies with deeper pockets for infrastructure can do the same. Now Redis Inc is trying to save their broken by design business model by changing the license. This casts a whole lot of doubt on the future utility and licensing of the Redis project. And this is what the community balks at.

[ufocia]

Who is the community?

[stephenr]

You keep making comments about this, as if Redis was build from scratch by the company that is now making it closed source.

They bought an open source project, and now that the original founder has stepped away they're trying to squeeze it for all they can.

The "big corps" that you claim are using it to "pirate software in a legal way" (a) have been contributing to the formerly open source redis project, and (b) are now specifically forking it to keep maintaining it as open source.

[pjmlp]

Doesn't matter, they are the rigthfull owners of Redis and the author has freely given ownership to them, and has been paid for.

Supermarket bills cannot be paid with pull requests.

[AnonymousPlanet]

Supermarket bills don't get paid by broken business models either. If Redis Inc never existed, Redis the software wouldn't be much worse for it. I'm starting to wonder who the entitled is in the first place.

[ufocia]

It's only broken when they go out of business. Just because you don't like the business model, doesn't mean it's broken.

[pritambarhate]

> rigthfull owners of Redis and the author has freely given ownership to them

By using BSD license Antirez has freely given it to the whole world, not the name Redis but the code. No matter how big the corporations, the cloud providers are just using that code the way Antirez intended when he used the BSD license. You can't blame the cloud providers for that.

> Supermarket bills cannot be paid with pull requests.

But one can become famous by writing quality open source software and this fame can be used to get very high paying jobs.

[stephenr]

> Supermarket bills cannot be paid with pull requests.

Nor with increasingly unnecessary and niche features aimed at "enterprise" customers, it seems.

One could probably even argue that buying the rights to the name of a popular permissively licensed project is a terrible way to pay said bills.

[ufocia]

One can argue a lot of things, and that's what we're doing here.

How is it terrible?

[stephenr]

It's apparently terrible because it didn't work.

[zelphirkalt]

> the whole free software movement

Eh no. What an overly broad generalization to read. Whether it is enough to make a living is another question, but that does not mean one must paint all of the communities the same color.

[pjmlp]

The fact that after 20 years this has become almost a daily discussion theme speaks for itself.

[AnonymousPlanet]

The problem is companies externalising development work on the boring parts of their software as "community edtions" and the like. That is a very distinct category of open source project and the only one that any of these discussions revolve around.

You seem to believe that all open source projects are in this category. That is not the case. You also seem to believe that there is always one company doing the most work and everyone else is just leeching off. That is also not the case.

[mort96]

I for one don't like it when companies do a bait-and-switch. It's fine to develop proprietary software, the problem is when you grow a user/customer base based on the fact that your software is open source and then turn it proprietary.

[stephenr]

With Redis it isn't even a case of "grow a user/customer base based on the fact that your software is open source and then turn it proprietary"

It's "buy the naming rights to an already popular piece of open source software and try to make a quick buck"

[ufocia]

Trust no one. Be self sufficient.

I, for one, will take the risk, reap the benefits and move on when factors are no longer conducive to my goals.

[sangnoir]

So I take it you endorse the Amazon-backed fork? Amazon too strives to be self-sufficient, and has moved on from Redis because the factors are no longer conducive to its goals.

[ufocia]

If it's legal, it's not piracy. It is merely availing oneself of an opportunity. If the authors meant to license the software differently, they should've done so.

I'm sure that (FL)OSS core/demo versions is not the ONLY sustainable way to make a living. There is no need for hyperboles.

You don't even need to author software to sustainably make a living. Don't limit yourself.

[lazyasciiart]

That doesn't seem like a very reasonable takeaway from an article which describes almost too many people announcing that they will take the last open-source version and keep developing it themselves for everyone else to use.

[palata]

If you only take, obviously there is no reason to complain. Now the problem is rather when contributors (those who "give", not those who "take") have to sign a CLA. Then the company who gets their copyright takes their work for free, to later use it in a non open-source project (assuming they changed the license, like Redis did).

I think it is valid to find this immoral. The solution is pretty simple though: do not contribute to open source projects that require you to sign a CLA.

[lukaszwojtow]

Using the code later in a non open-source project can happen also with MIT/Apache licensed code. Even without CLA. Does it mean that company that does it is immoral?

[palata]

If you use MIT/Apache code that doesn't enforce a CLA, the contributor keeps the copyright on their contributions. So no, that's not immoral, that's part of the license the contributor chose for their contribution (the contributor could make a PR and license their contribution with e.g. GPL: that would be their right).

What is considered immoral is to take the copyright from the contributors without giving any compensation.

[endisneigh]

No? They create a fork that maintains the existing terms. No cla required.

[palata]

That is not the problem with the CLA (of course you can fork). The problem with the CLA is that the company then uses the contributed code just like if it was their own, even though they did not pay for it.

Developers should be aware of that and, personally, I think contributors should never accept to sign a CLA. If the project requires a CLA, don't contribute.

[xandrius]

Yep, that's exactly it. Of course it makes sense: making requires several orders of magnitude more effort than using. But if a project changes/goes down, the community often just moves elsewhere, nothing major lost from their perspective.

And I think Open Source is based on the very few who decide to take it upon themselves to be the ones spearheading a specific project/task and share it with everyone else. Maybe it's not every single time me, sometimes it's you, sometimes it's Lucy or Mark, and that's how the roll keeps rolling for everyone.

So if a project goes down and nobody comes up to replace it, either it wasn't worth much or this is the time nobody took it upon themselves to do it (yet).

[jychang]

That's a dumb take. That completely ignores opportunity cost of such actions. You can't just spin up a fork like that; there's barriers to entry, network effects, etc which prevent that from being a simple solution.

[endisneigh]

You really can just spin up a fork

[LtWorf]

It's not "the community". It's "well funded startups".

People who use open source are very entitled. They'd be very angry also if the license was changed to GPL or AGPL.

I doubt most of this people have meaningful contributions to FOSS.