News for hackers. Stuff that matters.
Re:Undergraduate Upends a 40-Year-Old Data Science Conjecture by rollcat on Tuesday February 11, 2025 @05:30AM (#43011171)
In terms of practical engineering, this is also why I love to do side projects that reject existing bodies of libraries, and try to work up from first principles, and/or focus on composability rather than integration.

It's a trade-off, at first it takes longer to iterate on features, but sometimes a more minimal and/or composable tool finds its way to production. Real Systems are made of duct tape anyways.

There are no comments on this comment yet.