Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Redis changes license from BSD-3 to dual RSALv2+SSPLv1

github.com

Post reply on HN
Print view View original

21–25 of 25 posts

Re: Redis changes license from BSD-3 to dual RSALv2+SSPLv1

#21

Post by hgs3 »

blagie wrote: post #10
martinky24 wrote: post #9

Earlier quoted context omitted.

> I expect distributions like Debian, and Ubuntu by proxy, will move away from having a redis .deb as well. What about this change indicates that? This sounds like fearmongering, the new licenses really shouldn't affect this use case?

The new license is incompatible with Debian policy: https://www.debian.org/doc/debian-policy/ch-archive.html#the... The new licensing scheme effectively excludes redis from most of the open ecosystem, by cascading network effects from things like this. In this case, if you're building a package, if you pick redis, you've excluded yourself from distribution in many major distros. Ditto for many other similar effects.…

The "open ecosystem" does a poor job funding open source developers. It is completely understandable when a developer chooses to put food on the table over being part of their social club.

Re: Redis changes license from BSD-3 to dual RSALv2+SSPLv1

#22

Post by drewdevault »

evanmoran wrote: post #19
ChrisArchitect wrote: post #13

[dupe] ( https://news.ycombinator.com/item?id=39772562 ) A more recent development of interest: Redict is an independent, copyleft fork of Redis https://news.ycombinator.com/item?id=39789986

This fork is LGPL. Does anyone know of a fork that remains BSD? I’m sure this is fine for many people, but it isn’t ok for what I’m working on. Update: it looks like KeyDB is a good one. https://github.com/Snapchat/KeyDB

What are your concerns with LGPL?

Re: Redis changes license from BSD-3 to dual RSALv2+SSPLv1

#23

Post by rpeden »

ensignavenger wrote: post #3

Let the forking ensue. What a stupid move. They have illegally removed the BSD headers from contributions by various cloud vendors like Redhat and Amazon. I hope they can all get together to.maintain a fork, and maybe sue Redis Inc. for this unlawful infringment of there copyrights.

I don't see anything in the 3-clause BSD license indicating the copyright notice needs to be included in every file, though.

They've kept a copy of it in the repository, and it looks like they left the BSD header in files where the copyright holder's name was an external contributor.

They also mention that the license change only applies to future changes, not existing code - so it doesn't seem like there's any copyright infringement happening here, unless I'm missing something.

Re: Redis changes license from BSD-3 to dual RSALv2+SSPLv1

#24

Post by reconditerose »

evanmoran wrote: post #19
ChrisArchitect wrote: post #13

[dupe] ( https://news.ycombinator.com/item?id=39772562 ) A more recent development of interest: Redict is an independent, copyleft fork of Redis https://news.ycombinator.com/item?id=39789986

This fork is LGPL. Does anyone know of a fork that remains BSD? I’m sure this is fine for many people, but it isn’t ok for what I’m working on. Update: it looks like KeyDB is a good one. https://github.com/Snapchat/KeyDB

Several Redis maintainers and contributors moved to development here, https://github.com/madolson/placeholderkv. KeyDB is not bad though, it's an older fork that isn't very widely adopted.

Re: Redis changes license from BSD-3 to dual RSALv2+SSPLv1

#25

Post by blagie »

hgs3 wrote: post #21
blagie wrote: post #10

Earlier quoted context omitted.

The new license is incompatible with Debian policy: https://www.debian.org/doc/debian-policy/ch-archive.html#the... The new licensing scheme effectively excludes redis from most of the open ecosystem, by cascading network effects from things like this. In this case, if you're building a package, if you pick redis, you've excluded yourself from distribution in many major distros. Ditto for many other similar effects.…

The "open ecosystem" does a poor job funding open source developers. It is completely understandable when a developer chooses to put food on the table over being part of their social club.

The vast majority of the "open ecosystem" does a fine job funding open-source developers:

- In surveys, the vast majority of Linux maintainers are paid

- Both I, and many other people I know, have spent most of their adult careers being paid to do open-source

I have no problem with developers who "choose to put food on the table over being part of their social club." I have a serious problem with developers who pull a bait-and-switch, and:

- get support from the open ecosystem to build out market share and technology; and promptly

- pull a bait-and-switch, and try to milk their supporters for $$$

I think the point Redis is missing is that SaaS relies on **trust**. SaaS is almost always a better deal in the short-term, but once you're wedded to a vendor, you're relying on them to not extort you or hurt you. A SaaS vendor can potentially:

- discontinue a product with zero days notice

- totally ignore security, leading to all your data landing online

- raise prices 10x overnight, leaving you in an impossible situation

- "pivot" in all sorts of other ways which disappear your business

A major reason for doing things yourself is business continuity and risk management. Once a vendor has done something like this, it's a good sign they'll do it again. After a vendor pulls a trick like this, I wouldn't consider relying on them even for proprietary work I do.

(Footnote: This is also why I would never use Google's cloud or Oracle's cloud; there's a track record of business-killing moves).

Post reply on HN