Hacker News: minikomi

New comment by minikomi in "From zero to a RAG system: successes and failures"

minikomi — Fri, 27 Mar 2026 10:55:20 +0000

Yep. Semantically distinct and meaningful chunks wins every time over any kind of windowing or slice and dicing.

Unfortunately, many people are looking for a fire and forget solution over an existing rats nest of documentation debt..

New comment by minikomi in "A Japanese glossary of chopsticks faux pas (2022)"

minikomi — Fri, 20 Mar 2026 23:41:42 +0000

I think it's number 9 in the list

New comment by minikomi in "Lies I was told about collaborative editing, Part 2: Why we don't use Yjs"

minikomi — Mon, 16 Mar 2026 09:36:16 +0000

Component library page in the docs gives 404

New comment by minikomi in "Emacs and Vim in the Age of AI"

minikomi — Sat, 14 Mar 2026 09:30:08 +0000

I don't let it execute emacs lisp itself, but elisp generated in org mode babel blocks which is instantly executable is a fine way to have gptel improve itself.

( See "meta tooling" in https://poyo.co/note/20260202T150723/ )

New comment by minikomi in "Ask HN: How are people doing AI evals these days?"

minikomi — Wed, 11 Mar 2026 04:37:01 +0000

The more you can afford to build up your understanding of the problem space and define what inputs & outputs look like, the more flexible you can be with evals. Unfortunately, this is a lot of work and requires thinking and discussion with your team and those involved.

https://poyo.co/note/20260217T130137/

I wrote about general ideas I take towards simple single prompt features, but most of it is applicable to more involved agentic approaches too.

What I talk about when I talk about prompting

minikomi — Wed, 04 Mar 2026 16:54:27 +0000

Article URL: https://poyo.co/note/20260217T130137/

Comments URL: https://news.ycombinator.com/item?id=47250314

Points: 1

# Comments: 0

New comment by minikomi in "How I use Claude Code: Separation of planning and execution"

minikomi — Sun, 22 Feb 2026 03:54:58 +0000

I agree, and this is why I tend to use gptel in emacs for planning - the document is the conversation context, and can be edited and annotated as you like.

New comment by minikomi in "Evolving a Modular Dev Experience in Emacs"

minikomi — Sun, 22 Feb 2026 02:08:20 +0000

Given the recent surge of popularity in home-grown coding agents ala pi-coding-agent, I thought it would be nice to share my setup for semi-hands-on coding.

It's not a replacement for a more sophisitcated coding agent harnesses, rather an alternative interaction mode, and tools to support it.

Although the content runs fairly emacs-centric, consider it rather a nudge toward building your own small, self-owned tools.

Evolving a Modular Dev Experience in Emacs

minikomi — Sun, 22 Feb 2026 02:08:20 +0000

Article URL: https://poyo.co/note/20260202T150723/

Comments URL: https://news.ycombinator.com/item?id=47107385

Points: 2

# Comments: 1

New comment by minikomi in "What if writing tests was a joyful experience? (2023)"

minikomi — Fri, 06 Feb 2026 22:00:31 +0000

When writing Janet, i enjoy the judge[0] style of testing because it's so interactive. I added emacs helpers, one for running existing tests and one for updating the tests. It made for a nice repl-like experience, especially nice when writing grammars[1].

[0] https://github.com/ianthehenry/judge

[1] https://github.com/minikomi/advent-of-code/blob/d73e0b622b26...

New comment by minikomi in "Cowork: Claude Code for the rest of your work"

minikomi — Mon, 12 Jan 2026 23:42:28 +0000

i too could not resist

https://g.co/gemini/share/6aa102571d75

New comment by minikomi in "Show HN: Stop Claude Code from forgetting everything"

minikomi — Tue, 30 Dec 2025 09:10:53 +0000

I use gptel[0] with my denote[1] notes, and a tool that can search/retrieve tags/grep/create notes (in a specific sub folder). It's been good enough as a memory for me.

0: https://github.com/karthink/gptel

1: https://protesilaos.com/emacs/denote

New comment by minikomi in "Ask HN: Loneliness at 19, how to cope?"

minikomi — Mon, 29 Dec 2025 06:43:43 +0000

Group rides, coffee outside, bike packing are all amazing ways to make friends. Shared adventures however small make long lasting bonds.

New comment by minikomi in "The Illustrated Transformer"

minikomi — Tue, 23 Dec 2025 04:32:54 +0000

An example of why a basic understanding is helpful:

A common sentiment on HN is that LLMs generate too many comments in code.

For good reason -- comment sparsity improves code quality, due to the way causal transformers and positional encoding work. The model has learned that real, in-distribution code carries meaning in structure, naming, and control flow, not dense commentary. Fewer comments keep next-token prediction closer to the statistical shape of the code it was trained on.

Comments aren’t a free scratchpad. They inject natural-language tokens into the context window, compete for attention, and bias generation toward explanation rather than implementation, increasing drift over longer spans.

The solution to comment spam isn’t post-processing. It’s keeping generation in-distribution. Less commentary forces intent into the code itself, producing outputs that better match how code is written in the wild, and forcing the model into more realistic context avenues.

New comment by minikomi in "Show HN: HN Wrapped 2025 - an LLM reviews your year on HN"

minikomi — Sun, 21 Dec 2025 01:54:42 +0000

    1.▲
    Show HN: A songwriting DAW built entirely inside an Org-mode buffer(emacs.org)
    432 points | 2 hours ago | 89 comments
    2.▲
    Why I'm still using Soulseek to trade prompt-engineered MIDI files
    156 points | 4 hours ago | 42 comments

I admit I hastily tried clicking before realizing these were fake

New comment by minikomi in "2002: Last.fm and Audioscrobbler Herald the Social Web"

minikomi — Mon, 15 Dec 2025 07:29:38 +0000

Fond memories of browsing my downloaders on soulseek

New comment by minikomi in "DOOM could have had PC Speaker Music"

minikomi — Wed, 03 Dec 2025 05:06:44 +0000

Fascinating. It sounds awesome!

https://www.youtube.com/watch?v=havf3yw0qyw

New comment by minikomi in "How good engineers write bad code at big companies"

minikomi — Sat, 29 Nov 2025 00:10:33 +0000

So it's a problem of motivation you say

New comment by minikomi in "AI World Clocks"

minikomi — Sat, 15 Nov 2025 00:34:31 +0000

I find it a surprisingly similar mindset to songwriting, a lot of local maxima searching and spaghetti flinging. Sometime you hit a good groove and explore it.

New comment by minikomi in "Show HN: I'm making a detective game built on Wikipedia"

minikomi — Tue, 21 Oct 2025 06:30:27 +0000

Hey sometimes the "first letter" in the letter game when it's typed on the tape is in upper case .. kind of a hint :)