Hacker News: edg5000

New comment by edg5000 in "Flatpak – An Insecurity Nightmare"

edg5000 — Mon, 19 Oct 2026 08:18:42 +0000

Every time I tried Flatpak it did not work, always some other problem. Maybe this got fixed nowadays. I did not find it user friendly either, but this could be addressed by the distro. AppImage actually works and is simple, so I use that. Snap is trash because autoupdate cannot be disabled without code changes (what?!) last time I checked.

New comment by edg5000 in "Muse Spark Open"

edg5000 — Fri, 16 Oct 2026 04:48:36 +0000

Why this model split? And are your work volumes so high that you prefer these cheaper models over a subscription or are you simply trying to avoid the 200/m? (which will only go up).

New comment by edg5000 in "Ask HN: CAD for wood working with LLM support?"

edg5000 — Thu, 15 Oct 2026 09:47:50 +0000

In my experience the MCP is not needed, it just runs headless FreeCAD using CLI. The MCP can be useful if you want to run FreeCAD on a VPS, and then use ChatGPT e.g. on your phone (optionally via voice). But for local I'd go with the CLI. But they'll do just fine with the MCP, which probably wraps the Python API anyway (ideally, otherwise it could lack some features).

New comment by edg5000 in "Ask HN: CAD for wood working with LLM support?"

edg5000 — Thu, 15 Oct 2026 09:45:50 +0000

Astra in Codex CLI is good with FreeCAD if you let it use the FreeCAD Python runtime in a headless session. It is good at relying purely on natural language to design and revise. To follow specific CAD techniques you need additional prompting. Highly recommend. Uses quite a lot of tokens.

New comment by edg5000 in "Iosevka is an open-source, sans-serif and slab-serif monospace font"

edg5000 — Thu, 15 Oct 2026 04:07:59 +0000

True, it's not crossed. Which would you prefer over droid sans mono?

New comment by edg5000 in "Iosevka is an open-source, sans-serif and slab-serif, monospace font"

edg5000 — Wed, 14 Oct 2026 14:23:58 +0000

I've been using Droid Sans Mono for everything mono for a long time and can recommend it. But this font might be great too. Any Droid Sans Mono lovers out there?

New comment by edg5000 in "China Overtakes U.S. For First Time as Top Spender on R&D in 2024"

edg5000 — Tue, 13 Oct 2026 03:47:11 +0000

Shambles?

New comment by edg5000 in "Running the unmodified Linux amdgpu driver on an iPad"

edg5000 — Mon, 12 Oct 2026 09:13:16 +0000

For people who prefer a touchscreen input over a laptop, and if they'd want to do agentic work, or perhaps text to speech, image generation, music production, fancy spellcheck while traveling?

New comment by edg5000 in "3rd Yandex Cloud Data Center Was Hit"

edg5000 — Mon, 12 Oct 2026 09:08:13 +0000

I have to be careful with hyperboles, seems like my comment got flagged. I meant "the things I cannot find with Google I can find with Yandex". I actually have no idea about overall censorship, so "zero censorship" is a bit insane of me to proclaim in retrospect.

New comment by edg5000 in "3rd Yandex Cloud Data Center Was Hit"

edg5000 — Mon, 12 Oct 2026 09:05:50 +0000

I'm glad to hear that

New comment by edg5000 in "Total Annihilation, rebuilt for modern computers"

edg5000 — Mon, 12 Oct 2026 03:41:57 +0000

Zero Hour is played quite a lot competitively, but the scale is tiny versus SC2. It is quite balanced actually.

New comment by edg5000 in "3rd Yandex Cloud Data Center Was Hit"

edg5000 — Mon, 12 Oct 2026 03:38:59 +0000

I guess what I meant was "I can always find what I want", I have not tried searching for things that the Russian regime may find sensitive. So the better term would be "censored very differently", so it fills up Google's blind spots nicely.

New comment by edg5000 in "3rd Yandex Cloud Data Center Was Hit"

edg5000 — Mon, 12 Oct 2026 03:36:04 +0000

Ukraine has suffered much more material damage and loss of large swaths of land, and may have lost more civilians. Russia did lose a lot of troops, but their population is also larger. Russia did experience a brain drain, but Ukraine as well. I find "beating Russia" an odd take.

New comment by edg5000 in "500B Tokens Later: Letting AI Agents Decompile a First-Person Shooter"

edg5000 — Sun, 11 Oct 2026 09:30:53 +0000

> does still cover marginal inference cost Simply comparing to the larger models on OpenRouter implies that the pure hosting costs (equipment + power + minimal overhead) still exceed plan pricing if we assume all users always use their full weekly allowance.

So my conclusion is that it only works because the majority of users doen't fully utilize their allowance. Last month I used almost nothing of my Claude 20x plan (did use Codex though).

New comment by edg5000 in "OpenSCAD the Programmers Solid 3D CAD Modeller"

edg5000 — Sun, 11 Oct 2026 04:19:35 +0000

I found Astra to be surprisingly good at writing Python that generates FreeCAD documents. Arguably this says more about Astra than the ergonomy of writing FreeCAD XML, but because of Python the resulting code is very clean, because it puts anything verbose into functions. What could also work is direct OCCT from Python, haven't tried that.

New comment by edg5000 in "500B Tokens Later: Letting AI Agents Decompile a First-Person Shooter"

edg5000 — Sun, 11 Oct 2026 03:52:35 +0000

> subsidizing but not losing money

????

New comment by edg5000 in "500B tokens later: Letting AI agents decompile a first-person shooter"

edg5000 — Sun, 11 Oct 2026 03:50:14 +0000

I wonder about this too. Although, cracking was always possible. So what changed is that people can now modify, improve proprietary software with zero technical knowledge. I can now tell Astra or Fable (maybe not Fable because of safeguards?) to add a feature to Adobe Photoshop and it might actually succeed even when prompting purely on a functional level (e.g. the functionality I want out of the program). It would take days, but I think it would succeed. Call it PhotoBench?

New comment by edg5000 in "500B Tokens Later: Letting AI Agents Decompile a First-Person Shooter"

edg5000 — Sun, 11 Oct 2026 03:41:15 +0000

I sense the approach overcomplicates things. I wonder how long this would have taken in a single session. Maybe this is actually a textbook example of something where subagents make sense, but when I first started LLMs I was often overcomplicating the workflow with all kinds of orchestration. Now I just use one agent, it better allows controlling the output even if the agent works slightly longer. Most time is spent by me writing prompts and reviewing work anyway (for me at least).

New comment by edg5000 in "Step 5 Preview, a 1M-context MoE from StepFun, shows up on OpenRouter"

edg5000 — Fri, 09 Oct 2026 09:06:53 +0000

> carefully pairing models with use cases

I've always wondered if there were niches that some models are better at. I use them for software, electrical and mechanical engineering plus other things. What are some notable subject-specific differences you found?

New comment by edg5000 in "Mistral Large 4"

edg5000 — Wed, 07 Oct 2026 08:24:30 +0000

Large 4?