My approach since I was in High School and started programming has always stayed the same. It's actually very rare to have to innovate, starting from scratch. So whatever problems I need to solve, my first step is always looking for information. And more often than not, the solution is right there. I don't spend much time at the keyboard. Most of my time is spent perusing books, articles, blog posts, and source code.…
I'm also curious what kind of projects you work on.
Tell HN: A realization I've had about working with AIs and building software
11–17 of 17 posts
Re: Tell HN: A realization I've had about working with AIs and building software
#12Re: Tell HN: A realization I've had about working with AIs and building software
#13I've used Copilot for the past six months or so, mostly at work. I've talked to multiple colleagues and I seem to have much better luck using it than others, as in are you sure we're talking about the same tool? kind of discussion. I'm not sure why I manage to get so much more mileage out of Copilot than them. I mostly throw entire reams of C++ compiler error messages while working on heavily templated, very intricat…
This is a great write up! Hahaha :) I'm curious what sort of products you work on at your company and personally if you want to share.
- Company network is melting down due to the CI load? Hack together in a hurry a Git proxy cache written in a hundred lines of Bash. It's still in production years after I've left that sysadmin position.
- Company is trying to meld together two dissimilar pieces of software with wildly incompatible message buses and it's going so poorly it's threatening the entire project? Take a few days to improvise a message bus translator with a highly unconventional architecture. I then spend two hours in a design review meeting with the lead architects, who then decide to go along with it because they've failed to prove it couldn't possibly work and had no other option. This singlehandedly brought the project back under control, but it was pure black magic located right in the most load-bearing component possible and they took years to replace my Play-Doh with something saner.
- Company wants to redo the entire CICD? Write a Jenkins plugin that exposes a Debian source repository as an organization folder. With a suitable Jenkinsfile, it essentially turns Jenkins into a bona fide Debian build server, complete with proper dependency tracking. Management decided to go with a more traditional solution (OBS), at the cost of maintaining two different orchestrators and sets of workers.
As for my personal work, most of it is just various open-source contributions and faffing about on random stuff that never sees the light of day, but some of the things that actually ended up in the wild are:
- A Ghidra extension that can export relocatable object files out of a program selection. The simplest way I can explain it is Mad Max, but with program bits instead of car parts.
- Porting the epsilon firmware for the NumWorks calculator to the HP Prime calculator (another calculator manufacturer) and sending a pull request to the epsilon repository.
I should probably make an effort to actually finish more personal projects, release them and write up about them, but being chaos incarnate is somewhat antithetical to that.
Re: Tell HN: A realization I've had about working with AIs and building software
#14Earlier quoted context omitted.
This is a great write up! Hahaha :) I'm curious what sort of products you work on at your company and personally if you want to share.
I'm currently an embedded software developer, but I'm basically the guy you call when you're all out of options regardless of the problem. I've done plenty of debugging paratrooper-style for other teams, but some of my most infamous acts done in a professional capacity are: - Company network is melting down due to the CI load? Hack together in a hurry a Git proxy cache written in a hundred lines of Bash. It's still i…
Re: Tell HN: A realization I've had about working with AIs and building software
#15reading this made me go back to a Douglas Englebart line that has been deep in my head: these tools are for "augmenting human intellect." i think it still holds water in my experience with LLMs. i searched it up and found his paper on the subject (what a surprise, i had no idea this existed).
https://www.dougengelbart.org/pubs/papers/scanned/Doug_Engel...
Re: Tell HN: A realization I've had about working with AIs and building software
#16> AI is an amplifier of what you are reading this made me go back to a Douglas Englebart line that has been deep in my head: these tools are for "augmenting human intellect." i think it still holds water in my experience with LLMs. i searched it up and found his paper on the subject (what a surprise, i had no idea this existed). https://www.dougengelbart.org/pubs/papers/scanned/Doug_Engel...