Show HN: Postgres-SQLite sync engine source code released under FSL
github.com
github.com
6 months ago we did a Show HN about PowerSync [1], a Postgres-SQLite sync layer for building local-first and offline-first apps. We received lots of feedback, including comments asking about self-hosting and open-source.
Today we are announcing a major overhaul: making the product self-hostable and releasing the core sync service source code under the FSL license. We figured hackers interested in local-first software would like to know about this.
Docs: https://docs.powersync.com/self-hosting/getting-started
Release announcement: https://www.powersync.com/blog/powersync-open-edition-releas...
Background on how we decided to go with Sentry’s FSL: https://www.powersync.com/blog/new-open-era-for-powersync
Thoughts on what this means for the local-first space: https://www.powersync.com/blog/what-powersync-open-edition-m...
If you have any feedback or questions for us, we’d like to hear it :)
- Kobie, Ralf, Conrad, Phillip and team
Not sure, why the HN crowd is so silent about this release, this is BIG news.
To add some hints, why this is worth looking into:
- the PowerSync service was extracted from a working SAAS product that has been up since 2011 allowing to build offline-first apps in a low-code app builder
- this means it is NOT a research project that has yet to see extensive production use and iron out all the bugs and stuff
- the team behind is humble and very resourceful, judging by the amount of open source SDKs that take a lot of effort to maintain
- it supports _PARTIAL_ replication, and this is one of the few (if not the only one) of all the available open source solutions for local-first apps
Always happy to see a new alternative for small off-line-capable apps available to the average indie-developer!
> this means it is NOT a research project
Yeah, we’re quite proud of the system - the bugs in sync systems can be particularly insidious and it took us a very long time (years) to iron them all out. So we're happy to make such a mature system freely available.
Also - some bugfixes introduce performance issues too, another NB consideration for any component in a data pipeline. So along with the maturity comes a highly performant system: we can work with hosted and self-hosted customers to scale to millions of concurrent connections.