@qqq @didibus if you want to really go down a (fun) rabbit-hole, I highly recommend this blog: http://www.dalnefre.com/wp/ Especially the posts on "Kernel".

Let's build a Lisp from scratch, including stuff like fexprs, but make it so every expression is an actor! What could possibly go wrong? 🙂

@tbaldridge : nope, enough digging for today

now need to find a DB where I can store 100MB machine learning training data

while cassandra doesn't like > 16MB and CockroachDB doesn't like > 1MB

100MB? Sqlite 😛

oh a blob store

100MB is not entire DB size; it's the size of a blob

I hear “medium data” is the new hotness

a blob goes from 100MB to 10GB

(think collections of text / images / wav files)

Sounds like a good use-case for a Linux filesystem. Nested folders of files will take you a long way.

lol

I think you said the exact same thing last time I asked

Atleast you'r consistent 🙂