Post

Doing some investigation for the analysis side. Plan for now: - new repo for data analysis - run scraper locally - upload raw css to some storage (suggestions welcome) - analyze css and store results in duckdb - commit db files to repo - commit queries to repo Hoping to make analysis reproducible

Bart Veneman@veneman.dev

End of the year is approaching. I should probably start planning the next CSS Selection and not yolo it like I did this year. Does anyone have requests for what we should analyze?

Bart Veneman

@veneman.dev

@fubits.dev was it you who recommended to use duckdb?

Ilja

@fubits.dev

Hell yes - and I’d suggest to use @duckdb.org together with @marimo.io

29/9/26, 6:52 pm0 reposts3 likes

Ilja

@fubits.dev

you can also mix @marimo.io with FastAPI and deploy / expose multiple notebooks as both fully functional interactive notebook routes and as SSR’ed web app routes. That’s great for collaboration / preview / sharing.

Bart Veneman

@veneman.dev

That seems like a lot of new surface area to learn 😅

Ilja

@fubits.dev

Also see bsky.app/profile/tobi... DuckDB also has a Node client - for one project I’m using SvelteKit & load a persisted DuckDB file at boot time from remote storage. Using a hook I can refetch it from storage. Hence, I deploy the SvelteKit app once and update the data on demand. So many options!

Tobias Müller@tobilg.com

Deploy @duckdb.org to @cloudflare.social Workers via Emscripten and Rust/Wasm github.com/tobilg/duckd...

github.comGitHub - tobilg/duckdb-rust-workers: Deploy DuckDB to Cloudflare Workers via Rust/WasmDeploy DuckDB to Cloudflare Workers via Rust/Wasm. Contribute to tobilg/duckdb-rust-workers development by creating an account on GitHub.