> people who curate digital gardens of knowledge are in a unique position to be curators of focused streams of knowledge
I think you're onto something here. I save and organise articles and have been exploring ways to share them for some time.
There's two ways to organise knowledge - chronologically and topically. They're both important but for different audiences. Organising chronologically means followers can keep up with recent activity. It's simple and works and is the basis for RSS and social media feeds.
Organising by topic is for knowledge management where you can arrange and re-arrange content based on your understanding of the topic. One day you have a list of cool engineering articles then you might split that into articles about data etc.
I'm working on an app that tries to capture both. Each topic has a timeline of updates and each topic can be broken down into other topics.
I think a combination of chronological and topical organisation is best because each has their own audience.
Topics are great for visitors exploring your content for the first time and chronological updates are great for followers already familiar with your work.
This is the concept i’m exploring in my app, Kapa Notes. You write notes by topic and it generates timeline updates for you that you can edit as you like. I went with an outline based editor to reduce the friction of managing topics as much as possible.
I built a notes app specifically for capturing my HN links. It organises them both by date and topic so I can browse them by topic but also find “that one link from last week”.
I came to a similar conclusion about CRDTs after reading Figma's article.
I ended up implementing state machine replication using Cloudflare's Websockets and Redux and found the result to work quite well.
Redux actions seemed ideal for this problem because you can simply share the actions between peers. I couldn't find many redux-based solutions out there but it worked well for my use case.
I'm an Engineer with 7 years experience in a range of roles, stacks and companies. I've been Full Stack, Tech Lead and Engineering Manager, working with React, Node.js, Typescript, Next.js, SQL and AWS, in startups, scale-ups and small business.
I wanted incremental refresh in Postgres as well and found that you can manage your own table to get something close.
Basically you create a regular table in place of a materialised one, only aggregate data newer than what's currently in the table then store the new aggregates in table. Repeat this an interval.
I don't have the startup part yet but here's my one-person stack with Postgres, Node and React deployed on AWS with CDK using RDS, Lambda, S3 and Cloudfront. It's 100% in the free tier.
I think I’ve built something similar in React. It looks like a nested list but each item can have multiple parents. It syncs to GitHub and is public, I use it to share my recipes.
I implemented the same solution with Redux actions and Cloudflare Durable Workers. https://emojis.cadell.dev