I am trying to fix something that @Jimmy Miller said on here. In particular
There doesn't seem to be a way of building up higher level abstractions [on Observablehq.com] at the moment and saving and reusing them.
While is not technically true, because you can plug in a Firebase and build what you like, but that is waaay to involved for prototyping. Also once you did that nobody can fork your notebook and still get access to the backend, or they can, but the whole thing ends up inappropriately insecure. So the point stands that to do that well requires a fair amount of expertise. So I am fixing that.I am building a low config realtime multi-tenant state abstraction that supports end user forking and federated access. So you can build abstraction on Observable, but also so you can start to develop cross platform realtime workflows. In addition, anybody who uses that store can fork the containing notebook and still access their data, so it supports end user (re) programming. Furthermore it is a sane security model, that has per user isolation so forks can't steal access unrelated users data.I can talk about it now coz I can see the end and its totally possible! A forkable notebook realtime statestore is coming to Observable soon!https://futureofcoding.slack.com/archives/C03RR0W5DGC/p1659475446182609?thread_ts=1659196747.344019&cid=C03RR0W5DGC
🚀 Durable realtime state storage for Observablehq notebooks https://observablehq.com/@endpointservices/cellstoreI am most proud that end user forks keep data access available, so even for stateful notebooks you still retain remixability. I hope this unblocks a few use-cases.
2 weeks ago
Looks pretty awesome! Will definitely have to play with it. I like the setup of separating read and write access 🙂