clientdb VS workerd

Replicache (https://replicache.dev/) and clientdb (https://clientdb.dev/) are the only productized versions of this architecture I'm aware of (please do let me know if anyone is aware of others!).

But the architecture itself has been used successfully in a bunch of apps, most notable of which is probably Linear (https://linear.app/docs/offline-mode, I remember watching an early video of their founder explaining the architecture in more detail but I can't seem to find it anymore).

Basically the way authorization works is you define specific mutations that are supported and allowed, with a client-side and server-side implementation. The client side gets applied optimistically and then sync'ed and ran on the server, which applies authorization rules and detects and handles conflicts, which can result in client state getting rolled back. Replicache has a good writeup here: https://doc.replicache.dev/how-it-works#the-big-picture

This looks really cool, love seeing more innovation in this space!

At first glance this seems to be mostly targeted towards single-user apps where each user would have their own database that can be sync'ed to a remote server, but still isolated from data for other users, similar to the CouchDB+PouchDB model?

At least it looks that way since I couldn't see anything around authorization and conflict resolution. Not that there's anything wrong with focusing on this use case, a lot of apps can function perfectly fine this way.

A few other interesting new players:

https://replicache.dev/

https://clientdb.dev/

Standards bodies only standardize things after they've been proven to work. You can't standardize a new idea before offering it to the market. It's hard enough to get just one vendor to experiment with an idea (it literally took me years to convince everyone inside Cloudflare that we should build Durable Objects). Getting N competing vendors to agree on it -- before anything has been proven in the market -- is simply not possible.

But the Durable Objects API is not complicated and there's nothing stopping competing platforms from building a compatible product if they want. Much of the implementation is open source, even. In fact, if you build an app on DO but decide you don't want to host it on Cloudflare, you can self-host it on workerd:

https://github.com/cloudflare/workerd

In any case, I welcome this initiative with my open hands and look forward all the cool apps that people will now build with this!

[1] https://pyodide.org/

[2] https://github.com/cloudflare/workerd/blob/main/docs/pyodide...

[3] https://github.com/cloudflare/workerd/pull/1875

For ref:

- https://blog.cloudflare.com/workerd-open-source-workers-runt...

- https://github.com/cloudflare/workerd

workerd

I think this is for people who want to run their own cloudflare workers (sort of) and since nobody wants to run full node for that, they want a small runtime that just executes js/wasm in an isolated way. But I wonder why they don't tell me how I can be sure that this is safe or how it's safe. Surely I can't just trust them and it explicitly mentions that it still has file IO so clearly there is still work I need to do customize the isolation further. But then they don't show any info on that core usecase. But then that's probably because they don't really want you to use this to run it on your own, they are selling you on running things on their edge platform called "Wasmer Edge". So that's probably why this is so light on information.. the motivation isn't to get you to use this yourself, just to use this their hosted edge platform. But then I wonder why I wouldn't just use https://github.com/cloudflare/workerd which is also open source. Surely that is fast enough? If not then it should show some benchmarks?

i love how the main reference for workerd can be just one capnp file.

https://github.com/cloudflare/workerd/blob/main/src/workerd/...

this changed my world how i think about computing on the web.

if there was just a good enough js library as for lua and you could directly send capnp messages to workerd instead of always going through files. I guess one day i have to relearn c++ and understand how the internals actually work.

A significant chunk of it is open source: https://github.com/cloudflare/workerd/

Some of the configuration files for applications wind up being an entire language unto themselves, e.g., https://github.com/cloudflare/workerd/blob/1b5057f2bfcfedf146f6f79ff04e99903d55412b/src/workerd/io/compatibility-date.capnp

Edge runtimes are very good alternatives to containers that shouldn't be dismissed for "not being containers". They're often faster, more scalable, and cheaper than containers. Them being so lightweight also enable a "nanoservice architecture" – being able to run every service on a single computer instead of running different services on different computers and having to deal with network latency and unreliability.