Dad, startup guy. Ideas matter. Design matters. It's about we, not me.
hachyderm.io/@pevohr
Paul Rohr
Loading...
Bonus points to anyone who figures out clever ways to efficiently grab a time-limited snapshot (rather than a more global view)
the Atmosphere needs a LOT of care to be a functioning space for us all to have nice things.
there are many people here with experience in tradbiz, both sides of grant and philanthropic funding
there was even an expert in co-op structures at #atmosphereconf!
ask for help!
offer to help!
dms open!
The trick is to efficiently hydrate that postgres cache of selected metadata. How?
github.com/whyrusleepin...
Extracting a suitable snapshot from the network sounds like a job for tap/hydrant ... or maybe an early deliverable for hubble? 😉
cc: @ptr.pet, @bad-example.com
👀
The clustering impl (golang) + analysis files (csv, json) both look super-easy to read + use, but there are two missing dependencies:
1. the DATABASE_URL for a simple did/posts metadata cache (postgres) to query against
2. logic to calculate halfvec(512) embeddings for each post
