I found out about this via Andrew Kelley's blog[1], and I think he makes a great argument:
> In Why We Can't Have Nice Software, I point out this pattern of needless software churn in the mindless quest for profit. This is a perfect example occurring right now. Redict has already reached its peak; it does not need any more serious software development to occur. It does not need to pivot to AI. It can be maintained for decades to come with minimal effort. It can continue to provide a high amount of value for a low amount of labor. That's the entire point of software!
I hope that sufficient momentum/support can gather around Redict in order to maintain it for its existing users who aren't looking for a vector database, search database, generative AI widget, all in the one piece of software.
KeyDB is an existing fork that’s well supported and has a solid community for those interested. It takes a different philosophy to Redis but can be a drop in replacement in many cases
Interestingly, Drew DeVault, the creator of sourcehut, forks redis (into redict) in Codeberg, not in sr.ht:
> The source code is hosted on Codeberg, a Forgejo instance operated by a German non-profit, which should provide a comfortable and familiar user experience for anyone comfortable with the GitHub-based community of Redis® OSS.
> the Lesser GNU General Public license, LGPL-3.0-only.
> we will not be making use of any sort of Contributor License Agreement which gives one entity special privileges with respect to the copyright and licensing of Redict in a manner similar to that employed by Redis®. As a consequence, the copyright over Redict is held in common by all contributors, who would all have to agree to a future change of license, making it next to impossible that a similar change of license is in Redict’s future.
It's a shame the FSFE's FLA (https://fsfe.org/activities/fla/fla.en.html) isn't used by more groups. It allows you to provide some rights to the organisation (e.g. suing for license violations on your behalf), while you retain all your rights.
> In Why We Can't Have Nice Software, I point out this pattern of needless software churn in the mindless quest for profit. This is a perfect example occurring right now. Redict has already reached its peak; it does not need any more serious software development to occur. It does not need to pivot to AI. It can be maintained for decades to come with minimal effort. It can continue to provide a high amount of value for a low amount of labor. That's the entire point of software!
I hope that sufficient momentum/support can gather around Redict in order to maintain it for its existing users who aren't looking for a vector database, search database, generative AI widget, all in the one piece of software.
[1] https://andrewkelley.me/post/redis-renamed-to-redict.html