physicles

You certainly can do this for code, and it’ll give you great answers. But I find that, for whatever reason, info that I consume this way has a much shorter TTL in my brain than something I wrote myself.

I’ve been experimenting with writing a living spec — API endpoints, schemas, how things work at a low enough level that the LLM can fill in the details - and that’s been working well.

It also forces me to do the hammock-driven development necessary to build the right thing, and as a byproduct it keeps the design in my head longer.

> AI agent on my phone I can long-press power button and give spoken commands ("add milk to the shopping list"). Uses private home inference machine, see below.

For this part, are you using a tool-calling harness like Pi or Codex? Or did you sort of hack it together with prompts that tell it what json to emit?

I just got the rest of the “application layer” for my setup working yesterday, with a small set of evals. Auto-filing voice notes is what’s next for me.

What stack do you use for that home inference? I’m rolling my own with gemma4, but even with Claude helping with the implementation it’s a lot of moving pieces to figure out.

I use backblaze with restic. Got about 300GB up there, fully encrypted, keys in 1Password.

Not publicly unfortunately. I still go back to visit from time to time, and some of the things I could write about would definitely get me harassed if the wrong person found it. That's self-censorship for you.

TL;DR: it was far better than the rest of the world in 2020, marginally better in 2021, but 2022 was a shit show. I still remember Dec 7 as the day zero covid ended.

It absolutely is primarily about preventing people from learning about the world. You can’t run a system like China’s without controlling information. During the political meetings, they ramp up their blocking, and then they usually turn it down again after.

During the World Cup (2022, still zero covid in China), they started blurring out the crowd because people were noticing that nobody in the stands was wearing masks. They’d been told that the zero covid policy and draconian lockdowns were keeping them safe, and outside China was a Covid hell-hole. Protests followed four months later.

Nobody in China was told about western covid vaccines, much less allowed to get them.

VPNs aren’t widely used — there was a study posted here some months ago that estimated VPN use at around 1% of the population, too little to matter.

That’s why they’re a bit lax. They can afford to let some through to grease the wheels of commerce. You do have to be seriously unlucky to get into trouble.

(I was there during Covid, and some years before and after)

Oh god yes.

Fable 5.1 is a lot better than Fable 5 btw (edit: in terms of writing style). Not sure about opus 5.5 yet since I’ve only got one session in so far.

Oh wow, the Cirqa is exactly what I've been looking for: a wrist sensor without subscription and without a screen.

I had a FitBit Charge 4 some years ago, but it stopped working after a couple years.

Then I got a RingConn 2 last year, but like TFA I've decided that I don't like the ring form factor.

They really were separate. BuzzFeed News won a Pulitzer in 2021 (source: Wikipedia). It's unfortunate that they shared the same name.

> There was a moment when it dawned on him "I don't want to be anything like these people!".

Thanks for this great reminder that we can choose the kind of person we want to be.

Wow, I didn’t expect this kind of reply at all. What a labor of love. Can’t wait to try it out! What made you decide to emulate things at this level instead of using a more off the shelf hardware stack?

Amazing. This is nostalgia heroin for me.

How far does the HAL go? Screen 12 and 13, and writing directly to VRAM? WAIT for vsync? Call absolute for mouse usage?

Ha, Chris McCandless is to vanlife what Alex Honnold is to rock climbing: technically the same activity but with a very different level of consequence for failure.

I’ve been thinking about this a lot lately. For the last month I’ve been living in a camper van, working and traveling, which I’m really lucky to be able to do by choice. It’s lonely to be sure, but I also feel mentally far more resilient than I have in ages, and I’m trying to figure out why.

Being in nature is good for us, so that’s probably part of it, as is the constant practice of being self-reliant.

Another theory is friction. I can’t just roll out of bed to the couch, and back. I don’t have a fixed place to shower, so I swim in a river most days, and usually it’s freezing. Camping alone was terrifying at first, but objectively not that dangerous and not worth worrying about. All these little moments are a chance to exercise that mental muscle to do uncomfortable things.

Modern life is devoid of this friction. I could try to take a cold shower every day, but a not-cold one is just a turn of the wrist away. I could try to make myself go outside a bit every day, but I don’t need to.

This isn’t a new observation, but I can’t believe how stark the contrast is. I’m getting stuff done that I’ve been procrastinating on for years.

Jellyfin does all these things too, and it’s free.

Congrats on the traction you’ve got with this project. We used Gin back in 2016 for a while. I honestly don’t understand those Go devs who claim that the standard library is sufficient for building a serious API — there’s still so much boilerplate, and you end up writing your own mini-framework anyway.

The main reason we don’t use it anymore is that the default handlers don’t return the error type, which means error handling and HTTP status tracking get tricky and non-idiomatic as soon as your handlers start calling functions that can fail. It’s essentially a function coloring problem.

One could argue that HTTP status is a totally separate thing from normal errors, and I get that perspective, but that’s a significant ergonomic hit to take for the sake of philosophical purity.

I don’t know about the company’s ethics*, but the app is fine and it’s genuinely useful. For one thing, it has a map with the most popular running routes around you. And if you want to be part of an exercise community in the US or Europe (at least UK), it’s the only game in town. *Strava is one of the worst apps I’ve seen in terms of aggressively upselling you to a subscription. Genuinely infuriating.

Have you tried https://github.com/simulot/immich-go for dealing with the some of the complexity of google takeout? Not sure how much it actually handles, and whether it can be used easily if you're not migrating to immich, but it's what came to mind here. (I'm a new user of immich but never used google photos)

It’s not about ease of use and freemium, it’s the strong network economies at play, just like credit cards or social networks. It’s impossible for a competing product to get traction if it’s merely a little better than GitHub.

This is the real issue. We’re currently migrating from GitLab+Jira to GitHub. I did most of the migration, including our CI/CD with a self-hosted runner, and it was… fine.

The straw that broke the camel’s back was that Claude managed cloud agents are awkward to use with anything other than GitHub. In general, we realized that the whole world was on GitHub and we were swimming upstream.

But the original mistake was separating issue tracking and source control when we didn’t need all that power. Everything is so much simpler if you don’t have to maintain those integrations.

Maybe they’ll really enshittify GitHub in the future. We’ll cross that bridge when we come to it.

Nope, it was in the Midwest.

We did this on a train! Also StarCraft, also a coding competition, also felt like the coolest kids in the world.

I love that Naur paper! Also love this essay that extends it: https://hiringengineersbook.com/post/autonomy/

I literally maintain a conceptual-model.md that describes the “things” in our entire system, how they relate to each other, and some bits about how that maps to the code (the DB schema maps sort of closely to it but sometimes not that close, and so does the API, and that’s ok). I wrote most of the doc myself, but plan to coauthor with LLMs going forward.

So far it’s saved me tons of typing as I add features. I don’t have enough data yet on how well it will guide myself and others when adding stuff that really stretches the model, but I’m also optimistic that it’s a valuable approach.

Were you concerned about the overhead of shelling out to journalctl?

As someone who’s an experienced coder but sorta new to LLMs, I’ve learned that my intuition about what is and isn’t machine-readable for an LLM isn’t well-tuned. Better to do the benchmarking so you can be sure you’re not fooling yourself about whether use of YAML is warranted.

From the perspective of the “physics” of code (for lack of a better term), this feels right. We all know that if you have a generated asset, you basically mark that asset as read-only or you’re entering a world of pain. As more and more code is generated, the same forces will apply.

Harder for brownfield projects of course (and it basically replaces Jira, which would cost me all my social capital if I try), so I won’t be adopting it now. But I will be leaning into long-lived organic, grass-fed assets like conceptual-model.md that‘s critical for evolving the product forward but doesn’t strictly map to the types in the code.

I really enjoyed engaging with the article. Loved the humor sprinkled in, and the humility demonstrated throughout. Some great first-principles clear thinking here.

It’s almost as if we need to write some unit tests. For our profession.

Every now and then someone shares a small web link here (Kagi is one aggregator). It’s like survivors picking up the shards of civilization after the apocalypse. Of course such a project can remain viable and useful as long as it remains niche, which is virtually guaranteed as long as there’s no money in it.

It’s incredible how too much money corrupts everything it touches.

I switched to Linux in 2018, and I hear you on the quirks. But now, an LLM can fix basically any quirk you hit. I’ve been surprised multiple times. I’m also on Pop OS, which feels more batteries-included than Ubuntu.

I’d guess that’s what GP was referring to. I’ll also be a day 1 preorder.