The Web After Tomorrow (2015)

I just want to point out that if you toggle the night-mode switch at the top right of this blog, it doesn't do what you think it would do.

Discussed at the time: https://news.ycombinator.com/item?id=9765780

My initial reaction is: the fact that clients can talk directly to the database doesn't mean I necessarily want them to. Servers talking directly to the DB can be secured and the software running on them can be controlled. Clients aren't like that.

For the last 4 years I made a client side databases which solves this problem. You subscribe to queries and directly update the dom whenever something changes that affects the queries result. You can mix this with a client-server replication (couchdb or GraphQL) and voila, you have a website that always shows the correct data at each point in time. It even works offline, you have no loading spinners once the replication is in sync. Errors are handled in the background.

https://github.com/pubkey/rxdb

Is there a UI/color science reason why the author chooses a yellow background?

Interesting. Clojure and Datomic are at the heart of Roam Research... which definitely qualifies as a "web after tomorrow" effort.

I was amazed by how well a fully p2p solution fits right into the real-time data problem the author is describing.