This project hasn't been updated for half a year. Is it no longer being maintained?

[lllwan]

This project hasn't been updated for half a year. Is it no longer being maintained? If so, that would be such a pity.

[MiPnamic]

As per the Official Discord channel a release will kick-in soon.

[ashokgelal]

Unfortunately, it looks like this project has been abandoned. There has been many critical issues that haven’t been addressed at all for almost 6 months. The author, as much as he is helpful and nice, has been busy with other projects and no new commits have been made to any branches in this project. He has been promising new updates since February of this year but hasn’t follow through. At the very least I wish some bugs were fixed before archiving for many of us who invested in it from the beginning.

[franciscojavierarceo]

Maintaining open source is hard and Alex has done a lot for the community. We should keep the conversation constructive.

[sarahec]

@ashokgelal why not fork the repo, fix what you need, then file an issue and matching PR here? If you're using this in a commercial product, surely you could afford to contribute back.

[ashokgelal]

@sarahec Didn't mean this to say in bad faith. I was just mentioning that this seems abandoned with promise being made back in February that new updates were coming and that the author is being busy with other work. No, I'm not using this project and don't plan to (I was just doing some research and I came here after reading some posts on Reddit and HackerNews. Again, didn't mean to say it in a mean way just stating the obvious. And for the PRs, there are 21 open PRs pending (not from me, yes, but there are still pending PRs). Again, just want to be clear - I didn't mean to say this in bad faith, I was just wishing that if this was abandoned it would have been nice for it to me mentioned (and archived) for others to know whether to use this project or not.

[asg017]

To be fair, he isn't completely wrong — I've been saying that I'll publish new updates every few weeks, then end up not doing it.

sqlite-vec isn't abandoned. I still intend to push update and maintain the project, but I only have time on the weekends to look after it now, since sponsorships have dried up.

However, the extension still works as advertised, and I'm not aware of any bugs that effect correctness or that cause crashes. The bugs I haven't fixed have to do mostly with published artifacts, or annoyances like #54 . If there are any bugs that causes crashes/give incorrect results, please let me know.

I know I've said this before, but to reiterate: I still plan to maintain and give updates on this project very soon. Like a quick v0.0.7 release to fix some low hanging fruit, and merging some feature PRs soon after. I can't give good estimates yet, but sqlite-vec is far from dead.

[asg017]

@ashokgelal no worries, didn't read it as bad faith, it's hard to convey intent solely in internet comments

[ashokgelal]

@asg017 thank you! Sorry if it came out in a bad way, didn’t mean to. I appreciate your work even if I am not using this. I am following you on Discord and it was good to see you being still active and helpful. Yes, the delete one seems to be the one that is affecting many people (and what made me stayed away from it; I have always wanted to use this in a project ever since your posted it on HN). I hope you land up with more sponsors very soon. Maybe be even dual license it (but make it affordable for small devs like me who want to use it in a small product?)

[brankoradovanovic-mcom]

I'm personally quite happy with sqlite-vec as it is. Indeed, #54 is best described as an "annoyance", and the only thing that still bites me from time to time is #163, which should be fixed in 0.1.7 so yeah, a new release would be appreciated...

[dogada]

Is any comparison of sqlite-vec with other vector search solutions for SQLite? Turso, libsql (https://js.langchain.com/docs/integrations/vectorstores/libsql/), etc?

[asg017]

I haven't seen or created a good benchmark/comparison. I'll try to do that in the upcoming release

[vlasky]

Hi @asg017, really appreciate all the work you've put into sqlite-vec. It's good to hear it's still alive and evolving, even if weekend-paced. Totally understandable given your time constraints and the state of sponsorships.

I wanted to share something that's helped me maintain a few of my own open source projects under similar constraints. I've been using AI-assisted tools - most recently Claude Code - to take the edge off some of the heavier lifting, especially for implementing bugfixes, drafting new functionality, writing tests, and updating documentation. It hasn't replaced well thought out development, but it's been a helpful time saver.

If you haven’t explored tools like that yet, they might be worth a try. This could be an easy way to accelerate the release of some of the updates you're already planning.

Thanks again for keeping this project going. Looking forward to that next release when it’s ready.

[franciscojavierarceo]

@asg017 would be happy to help maintain the project if you're interested.

Alternatively, we could fork and maintain it as part of another large open source project I maintain, just a thought.

[francou]

sqlite-vec isn't abandoned. I still intend to push update and maintain the project, but I only have time on the weekends to look after it now, since sponsorships have dried up.

@asg017 I don't know if it could help the project in your situation but have you considered applying for an NGI grant ? The application is fast from what I've heard, and they give funds (50k€) quite easily to open source projects.

[SeanPedersen]

I really hope an update is coming this year (looking for freeing up space after deletions). If not is anyone aware of an actively maintained fork or similar project?