ppom
@ppom@mamot.fr  ·  activity timestamp last week

For the needs of #reaction, I wrote a very dumb #Rust#KV db which wraps BTreeMaps and a redis-like log file for persistance.
It's orders of magnitude faster than common KV #database (heed/lmdb, fjall, sled) because it's RAM-first.
But it doesn't have transactions, and it persists write operations asynchronously, after the write is done.
Those features make a lot of sense for reaction imo, but I don't know if it would be useful to others.

Question is: should I publish it as a separate crate?
ppom
@ppom@mamot.fr  ·  activity timestamp last week

For the needs of #reaction, I wrote a very dumb #Rust#KV db which wraps BTreeMaps and a redis-like log file for persistance.
It's orders of magnitude faster than common KV #database (heed/lmdb, fjall, sled) because it's RAM-first.
But it doesn't have transactions, and it persists write operations asynchronously, after the write is done.
Those features make a lot of sense for reaction imo, but I don't know if it would be useful to others.

Question is: should I publish it as a separate crate?
N-gated Hacker News
@ngate@mastodon.social  ·  activity timestamp 4 weeks ago

🚨 Breaking news: Supabase's new "Lethal Trifecta" feature ensures entire databases leak faster than a sieve 🏃‍♀️💨! With the innovative combo of LLM blunders, zero #security, and an express delivery system for #data breaches, it's a hacker's dream come true 😅🔓!
https://simonwillison.net/2025/Jul/6/supabase-mcp-lethal-trifecta/ #Supabase#Lethal#Trifecta #breach #database #hacker #news#HackerNews #ngated
argv minus one
@argv_minus_one@mastodon.sdf.org  ·  activity timestamp last month

I'm told that the purpose of #SQL was to allow people to query the #database without needing to concern themselves with how exactly the database works. Just ask it for what you want, and let it figure out how to make it happen.

Except…if the database isn't indexed appropriately for your query, it'll be slow, so you still do need to know how the database works. 🤔

I wonder if RDBMSes should automatically create indexes as needed? But there's probably a reason why they don't.