Relicensing CockroachDB • Hacker News Board

Re: Relicensing CockroachDB

#231

Post by paxys » Wed, Jun 05, 2019, 4:20 AM UTC

While I understand the move (similar to what Redis Labs, MongoDB, Confluent and many others have done recently), as one of the many contributors to these projects not affiliated with any company, it feels like a slap in the face. If such licenses gain prominence then having a healthy open source community for large-scale projects will be a thing of the past. As an outsider, it makes zero sense to be involved in the project if my code isn't really going to be open source until 3 years from now, and that can be delayed further or indefinitely if the company wants.

IMO the best way forward is for such projects to be run by non-profit foundations, not VC-backed companies.

Re: Relicensing CockroachDB

#232

Post by paxys » Wed, Jun 05, 2019, 4:28 AM UTC

dhuramas wrote: post #56

xwdv wrote: post #50

I’m still not sure why people open source great products then get outraged that someone else would simply resell the product and make money off of it, possibly even more than them. When I put something out as open source, I’m doing it fully aware that I am leaving money on the table, but it will be worth it if a community of maintainers springs up around the project and ends up making it far better and more useful th…

Does that mean we shouldn't revisit our decisions and all your decisions should be set in stone? CockroachDB tried something. It seems to not be working- so they are changing. It's not as if they are doing it retroactively.

Does everyone who has contributed to this project over the years have a say in this decision? My code itself will also be relicensed, whether I like it or not.

Re: Relicensing CockroachDB

#233

Post by reificator » Wed, Jun 05, 2019, 4:29 AM UTC

juped wrote: post #82

reificator wrote: post #30

Earlier quoted context omitted.

> Downvoters: if you think names for products do not matter you are mistaken. Disagreeing with your particular opinions on a particular name doesn't imply a belief that names don't matter.

You shouldn't downvote to disagree!

In fairness I didn't vote on that post, but I was assuming downvotes were for disagreement.

Re: Relicensing CockroachDB

#234

Post by paxys » Wed, Jun 05, 2019, 4:33 AM UTC

vortico wrote: post #215

NelsonMinar wrote: post #192

I'm curious about the legal process of relicensing. Did they have significant open source contributors from outsiders? If so, did they have copyright assignment? And if not, then how valuable is it to them to be open source in the first place? (Not trying to cause trouble; I admire Cockroach DB and its founders greatly! But I've watched open source struggle with this kind of thing for 20+ years and am increasingly cu…

If community patches are licensed under MIT or BSD, all CockroachDB has to do is include those license texts in their proprietary distribution, which is the only requirement by both licenses. Proprietary derivative works are allowed under MIT and BSD.

It's different in this case, since there isn't a separate proprietary distribution. The entire codebase is being relicensed.

Re: Relicensing CockroachDB

#235

Post by paxys » Wed, Jun 05, 2019, 4:34 AM UTC

vageli wrote: post #134

DVassallo wrote: post #44

Earlier quoted context omitted.

Stealing an open source project would be like saying stealing quick sort. You can't steal something if it was given to the public domain.

> Stealing an open source project would be like saying stealing quick sort. You can't steal something if it was given to the public domain. Most open source projects are not given to the public domain, but encumbered with clauses in the form of their license.

And none of the companies are breaching the license.

Re: Relicensing CockroachDB

#236

Post by reificator » Wed, Jun 05, 2019, 4:44 AM UTC

justin66 wrote: post #88

reificator wrote: post #30

Earlier quoted context omitted.

> Downvoters: if you think names for products do not matter you are mistaken. Disagreeing with your particular opinions on a particular name doesn't imply a belief that names don't matter.

The belief that names are particularly important to corporate software adoption and that CockroachDB is a good name for such a purpose is probably pretty rare.

I don't care one way or the other about the name. But if you're marketing directly to engineers who want a resilient solution, there's probably an argument to be made.

Re: Relicensing CockroachDB

#237

Post by didibus » Wed, Jun 05, 2019, 4:46 AM UTC

will4274 wrote: post #85

didibus wrote: post #63

Could someone say just take a normal open source license, like MIT, and just add a provision saying, if you are Amazon, Google, Microsoft, this license doesn't apply to you and you are not permitted to use the code in any way? Basically, just blacklist certain companies that compete with you directly.

The whole point of open source is that anybody can read, modify, and run the code. If you remove "anybody", it's just another proprietary license.

But the other option is that the license is made more restrictive for everybody. So at least that way, it is still open source for me, and a lot of other people, but not for direct mega competitor with too much leverage.

I know, it's a choice of two evils, and I don't know what's best.

Re: Relicensing CockroachDB

#238

Post by Hermitian909 » Wed, Jun 05, 2019, 4:47 AM UTC

paxys wrote: post #231

While I understand the move (similar to what Redis Labs, MongoDB, Confluent and many others have done recently), as one of the many contributors to these projects not affiliated with any company, it feels like a slap in the face. If such licenses gain prominence then having a healthy open source community for large-scale projects will be a thing of the past. As an outsider, it makes zero sense to be involved in the p…

I'm curious about this take. If I'm reading the article correctly it appears if you did not have plans to turn the software into a managed service this has no practical impact on you. Personally, for every open source project I've contributed to if this license has been in place rather than an open source one no one would have been impacted at a practical level (including those doing it professionally).

Aside from purely ideological concerns, what about these changes do you think will kill communities?

Re: Relicensing CockroachDB

#239

Post by didibus » Wed, Jun 05, 2019, 4:56 AM UTC

amenod wrote: post #68

didibus wrote: post #63

Could someone say just take a normal open source license, like MIT, and just add a provision saying, if you are Amazon, Google, Microsoft, this license doesn't apply to you and you are not permitted to use the code in any way? Basically, just blacklist certain companies that compete with you directly.

It's still not free/open/libre source anymore, because you are discriminating against someone, so it's not in-line with the set of freedoms as defined by both OSI and FSF. Whether this matters to you or not is another question, of course... :)

Hum, I feel there's some ambiguity there. Like FSF says from here https://www.gnu.org/philosophy/free-sw.en.html:

> A free program must offer the four freedoms to any user that obtains a copy of the software, provided the user has complied thus far with the conditions of the free license covering the software.

It's unclear what "obtains a copy" means. You could make it that certain companies are not allowed to "obtain a copy", and thus if they do, it is through illegitimate means.

For example, it goes to length to say Free software does not mean gratis. So I'd suspect it would claim that if software is sold, and you pirate the copy instead of paying for it, you have broken the conditions of the free license. But if you paid, and a copy is delivered to you, from that point on, the four set of freedoms are granted to you over the copy that was given.

So, I can see something where, it is gratis, but obtaining the copy is not granted to everyone, but only to some, depending on if you're a potential competitor or not. And that seems like it could still fall under that definition of Free Software.

Re: Relicensing CockroachDB

#240

Post by didibus » Wed, Jun 05, 2019, 4:58 AM UTC

toyg wrote: post #115

didibus wrote: post #60

I'm all for businesses coming up with whatever licensing terms they want. That said, as a user, it'll be a harder sell if I can't easily understand or validate what the license permits and doesn't. For example, with their new license, could I fork cockroachDB, rebrand it, add considerable improvements of my own, and then offer that as a service? Or what if I'm not offering cockroachDB as a service directly, but the s…

> could I fork cockroachDB, rebrand it, add considerable improvements of my own, and then offer that as a service? If you were to do that, even with current licenses, you probably couldn't call it CockroachDB anyway.

Ya, and I wouldn't, that's why I mentioned rebranding it.