charm makes some awesome libraries for building/serving CLIs over SSH: <a href="https://github.com/charmbracelet/wish" rel="nofollow">https://github.com/charmbracelet/wish
And this is why people aren’t going out of the loop any time soon. A normie asks the bot to make a project, it produces nonsensical complexity like this.
You're restating their point. Normies' output will still be bad compared to the output that good engineers would produce, with or without AI. You wouldn't want to use poorly built software regardless of origin.
Sure I would. I use it every day. Most software is crap, and it got worse once everything became a SaaS or embedded Chrome pretending to be a desktop app (or the worst, embedded Chrome loading a SaaS pretending to be a desktop app).
Hell, half of my use of LLMs these days is bypassing around crap in software written by people[0].
The metric in question is not how it's built. It's whether it solves a problem. Particularly for single or sub-100-users software, what matters is whether it does the job the user(s) like, the exact way they like it.
Speculating here, but maybe I'm also hitting at a disconnect between perspectives: I look at a product as tool to solve a problem. But some look at it as something to sell for money or Internet fame. For the former, distribution doesn't matter. For the latter, distribution is everything, and then, perceived code quality/effort signals actually start making sense.
TL;DR: I don't mind using shoddy software, but I agree it's a bit harder to sell obviously shoddy software.
--
[0] - <a href="https://news.ycombinator.com/item?id=49875994">https://news.ycombinator.com/item?id=49875994 I started calling this "Bypass-Driven Development".
HN data is accessible through the Firebase API: <a href="https://github.com/hackernews/api" rel="nofollow">https://github.com/hackernews/api
You don't need to as much as parse HTML to read it. Building a client by hand in C even should be pretty painless as long as you don't write your JSON parser yourself.
that's fair. thanks for pointing it out, there's a lot of helpful feedback here that will let me optimize this (admittedly) unnecessarily complex project :)
grim_io · · focus · HN ↗
Maybe your clanker can do it without.
tantalor · · focus · HN ↗
nateb2022 · · focus · HN ↗
neomantra · · focus · HN ↗
MisterMunchkin · · focus · HN ↗
TeMPOraL · · focus · HN ↗
nateb2022 · · focus · HN ↗
TeMPOraL · · focus · HN ↗
Hell, half of my use of LLMs these days is bypassing around crap in software written by people[0].
The metric in question is not how it's built. It's whether it solves a problem. Particularly for single or sub-100-users software, what matters is whether it does the job the user(s) like, the exact way they like it.
Speculating here, but maybe I'm also hitting at a disconnect between perspectives: I look at a product as tool to solve a problem. But some look at it as something to sell for money or Internet fame. For the former, distribution doesn't matter. For the latter, distribution is everything, and then, perceived code quality/effort signals actually start making sense.
TL;DR: I don't mind using shoddy software, but I agree it's a bit harder to sell obviously shoddy software.
--
[0] - <a href="https://news.ycombinator.com/item?id=49875994">https://news.ycombinator.com/item?id=49875994 I started calling this "Bypass-Driven Development".
dmos62 · · focus · HN ↗
nateb2022 · · focus · HN ↗
You don't need to as much as parse HTML to read it. Building a client by hand in C even should be pretty painless as long as you don't write your JSON parser yourself.
dmos62 · · focus · HN ↗
ahmd-sh · · focus · HN ↗