It's astonishing to me how difficult it (still) is to design a syncable local-first data model.
I keep thinking I've found a decent way, then realizing its flaws, then despondently noticing that the flaws were already discussed in Ink & Switch's article: inkandswitch.com/local-first.ht
Conversation
Abandoned the very general approach for my event sourcing-based model Flushout
github.com/saarw/flushout
Figured consistency requirements are app-specific and can be implemented on top with interceptors or even CRDTs, without incurring their impractical model size for other apps
1
7
This is a very graceful implementation, suggests event sourcing can be made general and compact. I'll see if I can adapt something like this to my purposes, where the data model is large and I must avoid grabbing a full snapshot except on first run.
Thanks! Sounds like you have interesting use cases! Realized models in my side-project apps aren't large enough to really need incremental updates, but get it for free with Flushout if I start saving command history to support undo or history views...