[adamwiggins]

You might want to read the original post--a core feature you're paying for is local-first sync, which is explicitly designed to solve a lot of the problems with iCloud.

I do want to see a world where there is a generic syncing service (maybe AWS can run one, but open standards / open source) similar to Dropbox or iCloud. I can pay one lump sum for all my storage and all my apps will connect through it. But so far no such thing exists outside of the crude file-based syncing of Dropbox, and that's not suitable for building realtime apps on top of.

[mantas]

Okay, then maybe that local-first sync can easily support multiple schemes and let old versions run without overhead to devs?

Personally I'm fed up with syncing both as a user and as a developer.

As a developer, I don't want to deal with infrastructure for an app. It's a massive headache to have 24/7/365 responsive system. I want to make apps, not be on-call sysadmin.

As a user, I don't want to worry who is going to sell my data after going bankrupt. And I'd prefer small dev shops don't waste their time on keeping network infrastructure up and running with security patches.

[adamwiggins]

> local-first sync can easily support multiple schemes and let old versions run without overhead to devs

I'd love that. Ink & Switch has done extensive research on how to enable this with p2p technology etc. Our industry isn't there yet... but lots of good folks are working on it. The Muse sync setup is a step in that direction.

> As a developer, I don't want to deal with infrastructure for an app. It's a massive headache to have 24/7/365 responsive system. I want to make apps, not be on-call sysadmin.

Oh yes. I spent many years carrying a pager for Heroku's infrastructure. Part of the appeal of local-first is the sync infrastructure is necessary to transmit data between devices, not for every single keystroke or gesture the user makes.