coffeeri

This is/was also my take. I’m skeptical that a probe-based network can be granular enough to reliably pinpoint a city, especially when some paths are much better connected than others (fewer hops, uncongested fiber, no throttling).

However, ipinfo still appears to rely on active probing to triangulate geolocation data, which suggests they believe these routing asymmetries can be modeled or averaged out in practice.

https://ipinfo.io/blog/ipinfos-probe-network

There is https://artificialanalysis.ai

Check the build.zig file. It defines the arguments for QEMU to run the OS.

You should be able to run zig build run.

This video [0] is relevant, though it actually supports your point - it shows Claude Code struggling with non-trivial tasks and needing significant hand-holding.

I suspect videos meeting your criteria are rare because most AI coding demos either cherry-pick simple problems or skip the messy reality of maintaining real codebases.

[0] https://www.youtube.com/watch?v=EL7Au1tzNxE

  Location: Cologne, Germany
  Remote: Yes
  Willing to relocate: No
  Technologies: Golang (Go), C, Python, NATS, PostgreSQL, Docker, AWS, QEMU, Linux, Dynamic Analysis, Reverse Engineering, Malware Analysis, Virtualization (Vagrant, Packer, Proxmox), NixOS, Microservices, REST APIs, Git
  Résumé/CV: https://www.linkedin.com/in/leander-kohler/
  Email: hn@leander.sh

M.Sc. Computer Science graduate with 5+ years student experience in cyber security research, Machine Learning and backend development. Strong background in malware analysis, reverse engineering, and distributed systems. Developed production microservices handling e-commerce operations. Multiple research projects at Fraunhofer FKIE focusing on dynamic analysis, unpacking, and emulation introspection (all with top grades). Maintain extensive homelab infrastructure using Proxmox and NixOS for personal projects and continuous learning. Looking primary for roles in cyber security, low-level development, or backend engineering. Available from May 2025 after a sabbatical.

Even better is https://simpletex.cn/

I am using NextDNS [0], which also integrates well within Tailscale across all my devices. Or are you looking for a solution that works offline within OpenWRT, without relying on third parties? It appears that there are AdBlock packages available for OpenWRT[1].

[0] https://nextdns.io [1] https://github.com/openwrt/packages/blob/master/net/adblock/...

Yes, it is marketed as a better “ls”. I was commenting regarding “maintained” in:

> A modern, maintained replacement for ls.

“Maintained” is referring to exa I think.

I think they are not referring to ls, but to exa [0], which is not maintained anymore.

[0]https://github.com/ogham/exa

It is the result of an energy dependence on Russia that has been built up over years, not due to any specific subsidiaries.

That comparison lacks as France heavily subsidizes nuclear energy as well. At the same time having to demolish and rebuild multiple reactors. Nuclear energy was never cheap. I don’t have the numbers but I doubt that France made the better net-deal.

The original video by Joe Grand: https://youtu.be/o5IySpAkThg

Interesting to see this pop up, especially after watching Jon Gjengset's stream yesterday on Implementing (parts of) git from scratch in Rust <https://www.youtube.com/watch?v=u0VotuGzD_w>

You don't need to do this online in many cases. For instance, the hash of WPA-secured Wi-Fi networks can be captured during the handshake of other devices.

You can bypass the templating in raw mode, by setting the request parameter `raw` to true.

https://github.com/jmorganca/ollama/blob/main/docs%2Fapi.md

Have you enabled the deep sleep state in the BIOS? It always depends what you are running on it. For me it's very light with only paperless-ngx and a smb share on a nixos. I also only have one RAM stick installed and set up powertop on startup.

I've also compared many 1L PCs from a performance/price/power consumption perspective. I decided on the HP prodesk/ elitedesk 400/800 G5 with the i5-9500T. It idles at ~2-3W and has considerably more computing and storage than these embedded solutions. I only see something like a raspberry pi be handy when relying on GPIO ports.

I love the cloud UX, but neither robot nor konsoleh. Hopefully they will migrate all services into cloud.

Depends on what you are trying to solve. If you don't like the fragmentations of dependencies on your system, this won't go away on a second machine. Also, I'd point you to hosted VS Code environments such as GitHubs code spaces or Gitpod.

You may use something that encapsulates your dev dependencies. Some use containers like docker, others use Nix (can be run on MacOS ). As Nix has quite a steep learning curve, there are abstractions such as devenv or devbox, which I haven't used.

Yes indeed, but it's not guaranteed that there'll be multiple clients connected to one server at a time, even if it's unlikely.

It is good that this does not work. As one IP might be shared by multiple accounts. A cache of the mapping IP --> AccountNo is also not favorable in terms of privacy.

I've had lately a look at QEMUs internals and saw their thread safe implementation of a hash table, capable of concurrent reads: qht [0].

If the author sees this, you might want to take a look at it.

[0] https://github.com/qemu/qemu/blob/master/util/qht.c

OpenAI seems to use Tiktoken [0]. It also covers GPT-4 token encoding.

[0] https://github.com/openai/tiktoken

If you are fine with self hosting a gitea instance, you'll be able to set up a pull mirror.

https://docs.gitea.io/en-us/repo-mirror/#pulling-from-a-remo...

You might have to adjust the minimal word length to 3.

I guess you are generally right. My prompts are lengthy, since I let GPT-3 handle the full html syntax of both sides, mathjax and so forth. I found the ada model to be not sufficient, but have not tried the other two models, I will try those! There is room for improvement on my side.

From a budget-conscious student's perspective, this amounts to roughly $20 per semester in my case, just for the tags. Despite this, I am still fascinated by the automation possibilities offered by GPT-3.

This statement is accurate for testing and experimentation purposes. Additionally, the $18 starting credit provided by OpenAI is a nice bonus.

I created a small Anki plugin that tags cards based on their content using the Text-Davinci-003 model. However, the cost quickly becomes prohibitive when a large number of tokens/ Anki cards need to be processed. To be fair that plugin is not working great, since tags should be viewed, merged and reused on a global scope.

If you're an Android user, you may be interested in Aegis [0,1]. It allows you to import data from various formats, either through a file or directly from another app like Authy (needs root access). It also allows for export options, such as encrypted JSON or unencrypted TXT/HTML. Additionally, it enables backups to file storage, I am using my Nextcloud as a target.

Overall, Aegis is my to-go open-source alternative to Authy.

[0] https://getaegis.app/

[1] https://github.com/beemdevelopment/Aegis

Without looking close at your suggestion, you might want to look at passage [0] by the creator of age. It's a fork of pass [1] using age as the backend.

[0] https://github.com/FiloSottile/passage [1] https://passwordstore.org

Heatpumps are great! Having a coefficient of performance equal to 4 (1kW electricity to 4kW heat) is possible today!

Unfortunately it is not as easy as just switching your gas heating with a heatpump. For one, heatpumps have a lower flow temperature, therefore a good insulation of the house is required. Also big radiators such as floor or wall are preferred.

Moving to heatpumps might start an entire house renovation which comes at a cost.