> be Justine Tunney
> start shipping software as a teenager
> spend years working on the part of computing everyone calls finished
> notice that binaries don't actually belong to one operating system
> leave a comfortable systems job to prove it
> build Cosmopolitan Libc around that idea
> compile once, producing one binary for Linux, Windows, macOS, and BSD
> no wrappers & no separate builds, just one artifact
> build redbean, a web server packaged inside a single zip file
> make Lisp run from a boot sector because the constraint was funny
> write a tiny x86 emulator because she needed one
> turn llama.cpp into a double-clickable executable
> make running an LLM possible without an entire software stack
> keep building things engineers normally dismiss as out of scope
> treat portability like a puzzle nobody bothered solving
Most of the patterns known as “good practices” in software development were created for humans that were expected to maintain/read that code.
That’s a thing of the past.
I for one want LLMs to solve coding, because every individual is an expert in their domain, with coding solved, everyone can use AI to build software for their domain.
The world is severely limited by coding.
I'm Miguel, 39.
DevOps + Data Eng from Mexico. Also content creation and mentorship.
Looking to connect with more engineers, startups and junior devs needing mentorship!
“We have a meaningful life when we do something that if we didn't do it, it wouldn't get done. That’s extremely meaningful and you should always aim for that”
— Peter Thiel
Fools learn from their own mistakes but wise people learn from other people's mistakes but learning from your own mistakes is quicker though mostly more expensive