At the time of writing (June 2025), the prevailing view in the software industry is that LLM-powered AI is either completely useless or will imminently destroy all software engineering jobs. As you might expect, the reality is somewhere in between. In this post, I’ll share my journey with LLM tooling, from reviewing an early, internal alpha of GitHub Copilot to my current daily usage of Cursor, ChatGPT, and the latest Copilot offerings. My perspective is that of a startup founder (of Workbr...| Mike McQuaid
Homebrew was the first open source project I’ve maintained where I’ve had to review and merge contributions from other users. Homebrew is also one of the most active community projects on GitHub with a consistently small team of maintainers (always under thirty in total, always under ten doing work every week). As a result I’ve had to figure out over the last twelve years how best to manage large numbers of contributions from users in pleasantly and efficiently for both maintainers and ...| Mike McQuaid
I’ve been encouraged by a mentor to think about what my core (engineering) values are (in the context of being recently promoted to be a “staff engineer” and having my eyes on being a “principal engineer” one day). This felt like something that could be of wider interest so here we go:| Mike McQuaid
I’ve worked on a few Ruby apps in my career at varying scales (Homebrew, AllTrails, GitHub, Workbrew) and there’s been a consistent theme: Ruby is great for moving fast (and breaking things).| Mike McQuaid
If you’re an open source maintainer lucky enough to have a significant number of contributors you need to learn to mentor efficiently. First timer issues are not the right good way to get people involved in your project nor mentoring individual first-time contributors. Instead, do things that help all of them.| Mike McQuaid