Skip to content
View cgwalters's full-sized avatar
👍
LGTM
👍
LGTM

Organizations

@CentOS @openshift @GNOME @coreos @projectatomic @ostreedev

Block or report cgwalters

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
cgwalters/README.md

Hi, I'm an engineer working on OpenShift 4, Fedora/RHEL CoreOS and as of recently focused on bootc.

I'm a big fan of Rust.

Other things I currently maintain include:

Online/social media

LLMs

See this blog.

I largely agree with this post by Graydon Hoare (the inventor of Rust). I had been looking forward to ending my career writing Rust, but...I am not typing it much anymore.

However, my record of maintaining FOSS for over 20+ years still stands. I commit to ensuring that all "core" software I write (i.e. software that may run on your computer, and I attach my name to it) and maintain will be reviewed by me. Further, I think it remains a baseline to have two humans involved, one author and one distinct reviewer (but now both may be assisted by LLMs).

I maintain software that is important, and I'm on the receiving end of often large LLM generated commits. I will treat your project with as much (or more) care as I do for my own.

However, if you get a PR from me, it is very likely that it is mostly LLM generated now, because as Graydon says:

I still write some code, but less and less, and more of it is around the margins: touchups, sketches of APIs and data structures, subtle stuff it's easy to be subtly-wrong about, or perhaps LLM-supervisory bits. Because the LLM really does often write the main logic as well as I would at this point, and faster, and more persistently.

My current experience is for deep, nontrivial work, I need to fix up somewhere between 10-15% of obvious LLM garbage (like reimplementing base64 encoding for no obvious reason) that still makes it pass subagent reviews etc. But, who knows what the future will be like.

An advantage for software engineers whose career in FOSS long predates LLMs like mine: you can easily see that I have the ability to write e.g. systems level Rust/Go/C etc. If I push a PR for a language/framework I don't know well, I will be crystal clear about that in the PR description.

Assisted-by/Generated-by

You may see this in my commit messages. Typically, I use OpenCode with a mixture of foundation models (Gemini and Claude). You can see my AGENTS.md and opencode config in my dotfiles repo.

Pinned Loading

  1. bootc-dev/bootc bootc-dev/bootc Public

    Boot and upgrade via container images

    Rust 2.1k 203

  2. composefs composefs Public

    Forked from composefs/composefs

    a file system for mounting container images

    C

  3. coreos/rpm-ostree coreos/rpm-ostree Public

    ⚛📦 Hybrid image/package system with atomic upgrades and package layering

    C 1k 217

  4. ostreedev/ostree ostreedev/ostree Public

    Operating system and container binary deployment and upgrades

    C 1.6k 355