Programmer by night. Also by day. Software Consultant at @MakerXAU. He/him.
Sydney, Australia
Joined October 2009
- Tweets1.4K
- Following550
- Followers447
- Likes539
Pinned Tweet
This week I did a bit of tidying and open sourced github.com/staff0rd/digitali…. Today the tidying (but probably not, and instead feature adding) continues.
Stafford Williams retweeted
New read! Our Principal Engineer, Stafford Williams, explores how to work with AI tools in a way that strengthens code quality rather than eroding it. In Steering the Vibe: Verify, he shares a practical mindset for tightening feedback loops, introducing verification as a first-class part of AI-assisted development. It’s a thoughtful take on building trust, consistency, and confidence in modern dev workflows, especially as AI becomes part of everyday engineering practice.
blog.makerx.com.au/steering-…
I starting writing a series of posts on finding the middle-ground between #vibecoding and professional #development. staffordwilliams.com/blog/20…
Stafford Williams retweeted
AI code assistants: game-changer or productivity killer?
Depends who you ask, and what phase of software you're in. In his latest blog, @staff0rd explores why feedback on AI coding tools is so divided, and how understanding the context of your build makes all the difference.
blog.makerx.com.au/ai-code-a…
Stafford Williams retweeted
Inspired by @staff0rd ' work on LLM agents, I built an MCP server that lets LLMs interact with #SEQ. It's amazing to see LLMs analyse system behavior by speaking SEQ's language through MCP! Check out here: ahmadreza.com/2025/02/buildi… #LLM #MCP @nblumhardt @datalust_seq
I compared @Netlify and Azure Static Web Apps in this #blog post and found surprising feature parity. staffordwilliams.com/blog/20…
Are you a #developer that needs local services exposed to the internet? Try @ngrokHQ staffordwilliams.com/blog/20… via @staff0rd
With a respectable collection of #docker containers I've started another #spacetraders adventure
staffordwilliams.com/devlog/…
Spaceships are once again beckoning and it's difficult to resist. staffordwilliams.com/blog/20…
In this post I demonstrate how to load and prompt large language models programatically. This approach can greatly reduce the need to write tedious parsing and processing code by leveraging the natural language processing capabilities of LLMs.
staffordwilliams.com/blog/20…
Even if you don't have a fancy GPU you can still run Stable Diffusion by using cloud hosts for as little as $0.40/hour
staffordwilliams.com/blog/20…
Using @vite_js with @capacitorjs makes #mobiledev fast and enjoyable - I #blogged about it here:
staffordwilliams.com/blog/20…
Stafford Williams retweeted
How to ship a feature faster:
Skip tests.
Skip types.
Skip docs.
Skip code reviews.
Skip error handling.
Cut corners on security.
Ignore performance.
Hard code values.
Ignore reuse. Just copy/paste.
Warning: This creates tech debt. This will slow you down in the long-term.
