dllu

11 時間前議論

> in Spanish, there wasn’t an English version, but it will do

I've just uploaded an English translation to: https://commons.wikimedia.org/wiki/File:Moyen_Orient_13e_si%...

If someone can proofread my translations that would be great.

dllu

28 日前議論

Yes I wrote the issue by hand and try to communicate well.

dllu

28 日前議論

I totally understand the point of view from maintainers. Review fatigue of low quality slop is a legitimate issue.

The worst ones are fully autonomous AI agents looking for open source projects and adding random pull requests.

But in some cases, I find a legit bug that needs fixing. For example, I want to get a particular program working in Wine/FEX on aarch64 [1], or I find a 12 second hang in Darktable [2]. The problem is that, as a software engineer working in a totally different discipline, I have no knowledge of the low level C code to fully understand what the problem even is, or how to fix it. All I want to do is to fix the issue and help other people avoid running into the same issue. Right now, on my machine, I maintain a set of custom patches to get everything working. But I am too dumb and ignorant to figure out how to create the fix by hand, so I can't submit a pull request (or when I do, I feel really bad about it. I honestly feel like a horrible person, e.g. when a project added a "No AI" policy soon after I submitted some AI-generated PRs [3]). Going forward, I feel like this sort of scenario is going to be way more common.

[1] https://github.com/FEX-Emu/FEX/issues/5512

[2] https://github.com/darktable-org/darktable/pull/21069

[3] https://github.com/FEX-Emu/FEX/commit/8c85096f98084ca9438b16...

dllu

先月議論

Yeah, if I take a dense grid of photos near my house, it would reveal a 500 m circle. But in practice I don't take _that_ many photos in the neighborhood. Also, the circle isn't perfectly centered on my home.

dllu

先月議論

For my personal website I have a lot of photography-oriented blog posts [1], but I have special code to strip out GPS info from the location if it's close to my home [2].

EDIT: my vibe-coding slop agent put my home GPS lat long in the example config in the README lol. Please don't rob my house; I'll go run git-filter-repo later.

[1] https://daniel.lawrence.lu/blog/2023-12-20-trip-to-europe/

[2] https://github.com/dllu/pupphoto/blob/main/gps.py#L81

dllu

先月議論

* Wikimedia Commons upload: https://daniel.lawrence.lu/blog/2026-03-25-uploading-to-wiki...

* Image viewer that can handle really big photos + run scripts via custom keybindings + CLIP search: https://daniel.lawrence.lu/blog/2025-10-22-sriv-simple-rust-...

dllu

先月議論

The NVIDIA DGX Spark that I use has a fairly decent ARM CPU that's roughly on par with the M3 Pro and AMD AI Max 395+ according to Geekbench, plus a GPU that's very powerful (RTX 5070 level). I'm sure it will be more than powerful enough for me to get a decent experience.

I also have a Macbook Air M2 running Asahi Linux.

dllu

先月議論

There's a Linux amd64 version and a Windows arm64 version. However, I use a Linux arm64 machine, and arm64 machines are going to be a lot more common going forward. I wonder if there are plans to release a Linux arm64 build?

dllu

先月議論

I also wrote some houseplant programming for my photo flow, which incidentally also includes Darktable [0].

As part of this, I vibe coded a script using AI to suggest captions and categories for photo uploads to Wikimedia Commons [1].

I also have my own image viewer [2].

[0] https://github.com/dllu/pupphoto

[1] https://daniel.lawrence.lu/blog/2026-03-25-uploading-to-wiki...

[2] https://daniel.lawrence.lu/blog/2025-10-22-sriv-simple-rust-...

dllu

2 か月前議論

Does screen sharing actually have reasonable resolution and latency befitting of the Apple Studio Monitor?

dllu

2 か月前議論

You can't have rotational symmetry with 5 pieces since that would require a piece in the center but the chess board has an even number sized.

dllu

2 か月前議論

Neat. Surprisingly, there are 388 solutions, and a lot of them look rather unintuitive.

    ........
    ...Q....
    ........
    ........
    .....Q..
    ........
    ........
    Q..B..Q.

    Q.......
    ........
    ........
    ........
    ..QQB..Q
    ........
    ........
    ........

My original intuition was to place the queens on unique rows and columns to cover as much as possible but it turns out there are solutions with three of them on the same row.

Python script: https://gist.github.com/dllu/698d5f71b2b9735c5c462ddf4a2f6fc...

Here's how it works:

0. precompute the attack patterns of each possible queen/bishop location as a bitmask, stored as an integer

1. generate candidate solutions, allowing attack rays to pass through other pieces, by brute forcing the positions of the 5 pieces and taking the bitwise OR of their attacks

2. out of the candidate solutions, check which ones are actually valid taking into account occlusion. Actually, you only need to check if the queen's horizontal attack is blocked by the bishop, as queens cannot block each other (the blocking queen herself has the same attacks so they effectively pass through each other).

dllu

2 か月前議論

Foreign investment isn't fake growth and money being spent in the country is definitely a good thing. It's how Singapore managed to kickstart its economy in the 1960s. Lee Kuan Yew tried very hard, and succeeded, in getting foreign corporations to set up shop in Singapore. The key is to capture value and move up the chain over time rather than getting stuck as a "cheaper back office".

dllu

2 か月前議論

I remember reading about this 4 years ago as the new Chris Lattner project and was super excited, though a little skeptical.

I think that nowadays with vibe/agentic coding, high performance Python-like languages become ever more important. Directly using AI agents to code, say, C++, is painful as the verbose nature of the language often causes the context window to explode.

dllu

2 か月前議論

I was thinking about doing the opposite for the common task of "SVG of a pelican riding a bike". Obviously, directly spitting out the SVG is gonna be bad. But image gen can produce a really stunning photorealistic image easily. Probably a good way to get an LLM to produce a decent bike-pelican SVG is to generate an image first and then get the model to trace it into an SVG. After all, few human beings can generate SVG works of art by just typing out numbers into Notepad. At the core of it, we still rely on looking at it and thinking about it as an image.

dllu

2 か月前議論

I vibe coded an image viewer for myself to address a big pain point that had been annoying me for years. [1]

[1] https://daniel.lawrence.lu/blog/2025-10-22-sriv-simple-rust-...

dllu

2 か月前議論

> to take part in a competition, election, etc., and try to win it

https://www.oxfordlearnersdictionaries.com/us/definition/eng...

dllu

2 か月前議論

Nowadays some parents went back to opting for cloth diapers. Apart from the obvious environmental aspect, there's the idea that ultra absorbent and comfy diapers disincentivize babies from signalling that they are about to poop. Apparently, babies can communicate when they need to go even quite early on, in what's called "elimination communication". This also makes them a lot easier to potty train later on.

[1] https://en.wikipedia.org/wiki/Elimination_communication

dllu

3 か月前議論

I vibe coded RSS feed for my static site generator a few months ago: https://github.com/dllu/dllup-rs/blob/main/src/main.rs#L850

and was pleasantly surprised to see my blog was already on this website!

dllu

6 か月前議論

That reminds me of Paquerette Down the Bunburrows [1] which is a very fun pathfinding game where the bunnies will pathfind to try to run away from you. It's not exactly what you described, but it is very fun and surprisingly deep and challenging.

[1] https://store.steampowered.com/app/1628610/Paquerette_Down_t...

投稿

A Handheld Battery-Powered Line Scan Camera

All about automotive lidar

Simple Rust Image Viewer

コメント