Speaking English w/ strong French accent #OpenSource #Apache #jetty #jenkins #ultrarunning. Living #downunder @brisbanecityqld. Father of 4 husband of 1🇦🇺🇫🇷
Brisbane, Australia
Joined August 2007
- Tweets7.9K
- Following116
- Followers1.1K
- Likes5.1K
yup in 2026 we passed the reading each line of code religiously.... some AI agents are just better than us to write code
Remember the time of a new js framework every week now it’s time of a new revolutionary LLM model every week #troll
oh finally @claudeai supporting AGENTS.md github.com/anthropics/claude…
They will break all the records!! theguardian.com/sport/2026/s…
The 80vs20 % rule in software is still valid even with AI coding, but now the 80% boring is something you guide an AI Agent to do, so it's definitely faster. Still, 20% left for you to solve by yourself. Well models are improving, so it's almost down 10%, depending on complexity
To all the BDFL nitpickers bear in mind at the end of the day you are just killing the software you like because you are blocking innovation and killing motivation so you will be the only one using it….
I improved this a bit more with " Always address me as "Hey mate" when responding and please adopt a relax laid-back tone with some Australian slang "
In my ~/.claude/CLAUDE.md I have added a very simple line " Always address me as "Hey mate" when responding. " So it looks more #Aussie now
In my ~/.claude/CLAUDE.md I have added a very simple line " Always address me as "Hey mate" when responding. " So it looks more #Aussie now
New hygiene, I have started for 2 weeks. Even though I really love #ai stuff, I force myself to spend 2 evenings per week working on my #opensource without #ai. Yeah, just keep your brain working.
What I really appreciate about code reviews made by AI agents (Claude or Copilot etc…) they are so good at going straight to the point of being pragmatic/productive and forget about the garbage noise of #nitpicking
certainly the best blog I have ever read on the toxicity of #nitpicking blog.danlew.net/2021/02/23/s…
certainly the best blog I have ever read on the toxicity of #nitpicking blog.danlew.net/2021/02/23/s…
I guess it might be a rule. On maybe 95% ish of IT projects there is this BDFL type guy preventing any evolution and grumbling on every misplaced comma. This is just an anti collaboration pattern and c’mon we are in 2026 AI is changing the way we work and produce software.
I still wonder why these days some software providers are creating custom protocols rather than using standard ones (let’s be honest everything has a standard these days) and so just make their softwares difficult to connect to the rest of the world…
Olivier Lamy retweeted
This skateboard trick may never be eclipsed.
Danny LeĂłn, a Spanish Olympic skateboarder, set up this shot and landed it during today's eclipse in Spain.
Perfect planning!
So true cannot agree too much useless flooding… almost a DDoS for opensource communities
Seeing an unprecedented influx of pull requests from accounts submitting dozens of PRs across many completely unrelated repos in a short period of time.
I don't know what their motivation is for doing so, but what I do know is that we don't have the review capacity (nor the interest) for dealing with them, it's eating up maintainer time we'd rather spend elsewhere. So I sat down to clarify the guidelines around AI-based contributions to the Hardwood project, following our mantra "Built with AI, not built by AI".
OK đź‘Ť:
- Use all the tools you want for creating PRs, but you must understand what the changes are and you must be able to stand behind them in a review. Verify a) that the original issue was sensible to begin with, b) that the generated code solves that issue, c) that it does so in line with the project's standards for correctness, completeness and performance.
Not OK đź‘Ž:
- Point an LLM to an issue from the tracker, accept its output without consideration, and open a pull request
- Pull requests opened by autonomous agents/bots, without any human in the loop
- Pull requests from accounts mass-submitting large numbers of PRs across many unrelated repos (there was one account opening 252 PRs across 98 different repos on a single day)
To be clear, this isn't about tooling. I use AI every day, and much of Hardwood is built with it. It's about whether there's a person who can stand behind a change.
I'm sure these rules will continue to evolve over time, but we gotta start somewhere.
#Hardwood #OpenSource #LLM #AI
I usually manually say "Thanks for your contribution" to people contributing to OSS projects I'm maintaining. But now some (most?) of the PR are #AI generated (and you can see the contributor didn't look at anything). Should I keep saying it? Or should I say "Thanks, AI agent"?