krschacht

Qwen 3.8 is surprisingly usable for a local model, especially inside the Claude Code TUI

In the last 18 months I tried a Framework twice and both times ended up returning it. I was running Arch both times and had too many low level issues.

I couldn’t get hibernation or sleep to work reliably. And about once a week I’d get a random freeze or crash that required me to reboot and lose all the windows I had open. I spent dozens of hours with Claude chasing things down, disabling various power management features, trying different kernels.

I really appreciate how much Apple computers just work. I’ve since invested heavily in Karabiner, Aerospace, Superkey, and a few other utilities to get close to the level of customization I had in hyprland. I still miss the polish I used to have, but I can close my laptop lid, walk out the door, and 100% trust I’ll open the lid and resume work. That counts for a lot.

I’m keeping an eye on Omarchy and Framework to see if they eventually solve all issues. Maybe in a couple years I’ll try again…

Which god are you asking about? People have come up with so many different ones. Do they all count?

I’ve been starting companies my whole career. My first real one was a software dev shop while in college, I sold it my senior year and dropped out. It wasn’t much of a financial success but it was more fun than typical college job. So sure, a success by my standards but not by most others.

My next one paid me a salary, but I had a falling out with my cofounder and sold my shares for very little to get away. Success? I felt like it, again, maybe externally.

Two duds after that, one I raised money for so I got paid a salary and I sold it to recoup most of investors money.

The one after that was a big hit: we grew to 50 employees, 8 figures in revenue and 9 figure exit.

I like to say: if I could find the perfect job, it’s much easier to work for someone else than start my own thing. But I have specific interests I want to chase so creating a company is my way to create the “perfect” job for myself. If you take this approach, it’s much more likely to be successful in your own eyes.

At the end of this article it states, "Our tests gave models the vulnerable function directly, often with contextual hints (e.g., "consider wraparound behavior"). A real autonomous discovery pipeline starts from a full codebase with no hints." I'm not a cybersecurity expert, but isn't 80% of the challenge finding where the exploit lives in the code!?

That really undermines the author's claims. This article feels dishonest in it's claim that "small, cheap, open-weights models ... recovered much of the same analysis."

Dang, that would be unfortunate if this is driven by ads.

I think I see the problem… and a possible solution.

I think the main problem with attempting to document this is that the system would not be running off of it. Your infrastructure document is automatically read and drives the deploy (or whatever). But if you want to make a change to a human’s responsibilities, you don’t get the simplicity of updating your organization documentation and clicking “execute.” So this new documentation you propose would always be lagging documentation rather than the actual driver of organizational behavior.

But! What if it was? What if all the managers in an organization were AI systems? They would read diff in the org chart and it initiated the communication to the respective human employees.

I could imagine testing this in a coffee-shop level business right now in which the LLM is probably capable of all the strategy and management decisions needed to effectively run it, operating within the constraints of policies and procedures all cleanly laid out in documentation.

For something like tests, where I have very specific opinions on how I want them written, I have a simple doc (tests.md) and I’ll regularly tag Claude with it.

Claude writes a bunch of new code and I’ll tell it, “Before I review this code, make sure all tests adhere to the guidance of @tests.md” (you can probably make this a slash command too)

I find that if I put these instructions in the system prompt, far down in a conversation that’s used lots of the context window, they will only loosely be followed. But when I tag it in like this, Claude will strongly and thoughtfully follow the guidance and examples I’ve written up about how I want my tests.

antirez — how do you reliably get Claude to re-read the file after compaction? It's easy to let Claude run for awhile, it compacts and starts getting much worse after compaction, and I don't always catch the moment of compaction to be able to tell it to re-read the notes file.

Run your own server for GitHub Actions. There is a simple library they have you install so your runner gets registered with your repo. Then you can SSH in whenever a job fails. This lets you fully inspect the state, and to execute one-off commands to test theories. It’s a much faster way to iterate.

I don’t lock down my kids computer use. I know risks exist, but I think their unrestricted access to computers and the internet is far more beneficial than harmful.

I know I’m highly unusual amongst my friends. I’ve also found it odd that the more knowledgeable someone is about tech, the more scared they are of their kids using the internet.

But just like riding a bike and swimming in a pool are extremely dangerous, yet I encourage my kids to do both of these things and instead just educate them about risks. Similarly, I think the benefit-vs-risk of the internet is FAR better than a bike & pool, so I just educate them.