News for hackers. Stuff that matters.

flurdy

Re:Gemini 4 Carbon on Wednesday October 28, 2026 @07:14PM (#50282306)
I only use the Google models via OpenRouter. Partially as setting up directly is a pain, and keeping the Gemini CLI working was another pain but also as a buffer to protect my Google account as you mention
Yeah, leave vulnerabilities exposed all weekend... ;p

I have been with clients that has that 'don't push to prod on Friday' like it was 2010. It is so frustrating, basically loosing 20% of your work week and batching work for a deluge on Monday.

It is a red flag of flaky architecture and low confidence in ops processes.

Though pushing at 5pm on a Friday is a bit of a dick move.

Re:I'm in a Meeting on Friday October 09, 2026 @09:47AM (#50020500)
https://hackertyper.com/ and others. Full screen code stream
Re:I'm in a Meeting on Friday October 09, 2026 @09:45AM (#50020475)
You try ignoring when the family helpdesk is called. You never hear the end of it.
Re:I'm in a Meeting on Friday October 09, 2026 @09:43AM (#50020444)
Yeah we need a dial to mix accents and dialects, and level of English fluency. I could want 50% English, maybe some Northern English, Welsh or Scottish, and mix in Indian accents, various American, Spanish, Portuguese, Polish, Scandinavian, the odd token Aussie. Etc.

That I could enable more South African, or Middle East accents in some meetings would be true to real life as well.

Not Finnish. The silence would not work :)

Re:GPT-6 Sol and Luna on Tuesday September 22, 2026 @07:34PM (#49809720)
Ah, poor Terra. Left on her own.
Re:The agents never stop. A council can on Tuesday September 22, 2026 @03:52AM (#49797911)
One more prompt. Knowing when to stop, when working together.
Re:Don't Let Anyone Take Away Your Big Box of Cables on Friday September 11, 2026 @04:27AM (#49655180)
I have also accepted I need duplicates of stuff for convenience.

I have one box that has all that I need when we go away for the weekend/week (by car, not plane). All the chargers since the family now has a lot of phones, headsets, kindles, cameras etc. 2 HDMI cables, audio splitters, speakers, extension leads etc that I nearly always need when away for a while.

And then my commute bag has a small essentials bag for the now rarer times I work elsewhere.

All with duplicates of what I already have at home. That I before instead used to unpack and reuse at home, then find and repack all the time, and forget some. Not a problem anymore. My tech packing is literally picking up a box and put it in the boot of the car. Done.

Re:Don't Let Anyone Take Away Your Big Box of Cables on Friday September 11, 2026 @04:19AM (#49655125)
The problem with the big box of cables is that they multiply over the years. I now have 3 big boxes of cables. And a few smaller ones.

At least they are see trough plastic so I can easily see what is inside them though.

Re:iPhone Duo on Wednesday September 09, 2026 @03:10PM (#49632408)
Yeah. A foldable that is the size of the 12 mini so that you can use it with one hand but still fold out to watch something etc.

But instead we got the largest screen before unfolding…

Re:Ask HN: How do you manage skills files? on Monday September 07, 2026 @05:24AM (#49596034)
I do both, or rather I do 'skill browsing', for new ideas to then evaluate the skill with my agent if they are useful. Most are not, but some I extract ideas from to augment my own skills https://github.com/flurdy/agent-skills/tree/main/skills#shar...

Though most of the time my skills are just things I found useful and could avoid repeating myself by having as a skill.

That I also use it to route model used with https://github.com/flurdy/pi-skill-model-router is also a reason

Re:The other Sean Byrne doesn't exist on Saturday August 15, 2026 @08:12AM (#49309957)
Now we know why Elon named his kid 'X Æ A-Xii'.

UUID names next? v7 so we get the age...

Re:Vinyl Cache and Varnish Cache on Thursday June 18, 2026 @11:07AM (#48586564)
Vinyl flooring etc is at least related to varnishing floors?
> There's an issue with the selected model (claude-fable-5). It may not exist or you may not have access to it. Run /model to pick a different model.
Re:Job: Head of Stonehenge on Tuesday June 09, 2026 @09:18AM (#48460768)
Ask one of the Ylvis brothers
Re:Anatomy of the .claude/ Folder on Friday March 27, 2026 @12:32PM (#47544869)
I was glad when linux went with the .config standard for most dotfiles.
That all random game and messaging sites now wants my kids' passport uploaded to some random 'id verification company' is madness.

But now instead, my 11 year old's Roblox thinks she is 18 because she wore glasses in their age verification webcam tool. And it can't be changed unless she uploads a passport, which I will never allow.

Please, gov.uk introduce a gov ID verification service? I could trust that, -ish, I have worked with public sector clients several times...

Only when chatting in a large channel at work, did I realise nearly 1/3 of the people there also set theirs as 1/1/1970. Which I presume is the first date that phisers will try to enter to reset people's accounts.

I am fully aware that my standard fake birthday is now used by me in some many places, that I have started to have a fake fake birhday. I should really just randomise and store it in my password manager.

But obviously the context of this OP story ruins all that.

Re:Anthropic Drops Flagship Safety Pledge on Wednesday February 25, 2026 @08:58AM (#47151525)
Many startups that build features which sit on top of Claude/ChatGPT/Codex, etc. And I think:

You are just one new feature announcement from Anthropic/OpenAI away from irrelevance.

Same as it was when people built their busineses on top of AWS a decade ago

Are people still reading PRs in detail manually?

I am after the automated PR agents have all passed a PR I tend to let Claude Code and Codex give me a summary, with an MCP skill to read the requirement story. I trust their ability to catch edge cases and typos more than me. I just check the general structure of the PR

It's the same reason why I have, for more than a decade, been so frustrated with people refusing to consider proper pair programming and even mob programming, as they view the need to keep people busy churning lines of code individually as the most important part of the company.

That multiple AI agents can now churn out those lines relatively nearly instantly, and yet project velocity does not go much faster, should start to make people aware that code generation is not actually the crucial cost in time taken to deliver software and projects.

I ranted recently that small mob teams with AI agents may be my view of ideal team setup: https://blog.flurdy.com/2026/02/mob-together-when-ai-joins-t...

Re:15 years later, Microsoft morged my diagram on Wednesday February 18, 2026 @05:53AM (#47059681)
LOL, calling Scott Hanselman a 'VP of something' is funny. Been listening to his stuff for years, even when I despised MS. Always seems genuinely nice. Probably one of the main reasons I these days have a more positive image of Microsoft.
Re:Agent Skills on Tuesday February 03, 2026 @10:29AM (#46872193)
It's why I wrapped my tiny skills repo with a script that softlink them into whichever is your skills folder, defaulting to Claude, but could be any other.

I treat my skills the same as I would write tiny bash scripts and fish functions in the days gone to simplify my life by writing 2 words instead of 2 sentences. Tiny improvement that only makes sense for a programmer at heart.

[1] https://github.com/flurdy/agent-skills

Re:London became the rest of the startup capital on Thursday January 29, 2026 @07:46AM (#46809464)
But it is also really good. I love the completely enclosed platforms, ie shielded from the track and train by a glass wall/doors, like the Jubilee line, but all the way to the ceiling. This makes it both safe and very quiet.

Though the platforms are huge, as the trains are long, you have to really make a conscious decision on which exit to use as they come up very far from each other. Unlike other tube stations, where if you don't pick the most optimal exit, you just have to cross the road.

I was so confused. Why is domain driven design especially good for debugging? I guess context is bound within the models... And then all the other comments were just talking about debugging tools. Glad I was not the only one.

Older »