Marc Schwieterman marcisme

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

A Possible Vision for Macros in Swift

As Swift evolves, it gains new language features and capabilities. There are different categories of features: some fill in gaps, taking existing syntax that is not permitted and giving it a semantics that fit well with the existing language, with features like conditional conformance or allowing existential values for protocols with Self or associated type requirements. Others introduce new capabilities or paradigms to the language, such as the addition of concurrency or comprehensive reflection.

There is another large category of language features that provide syntactic sugar to eliminate common boilerplate, taking something that can be written out in long-form and making it more concise. Such features don't technically add any expressive power to the language, because you can always write the long-form version, but their effect can be transformational if it enables use cases that would otherwise have been unwieldy. The synthesis of Codable conformances, for ex

True Color (24-bit) and italics with alacritty + tmux + vim (neovim)

This should make True Color (24-bit) and italics work in your tmux session and vim/neovim when using Alacritty (and should be compatible with any other terminal emulator, including Kitty).

Testing colors

Running this script should look the same in tmux as without.

curl -s https://gist.githubusercontent.com/lifepillar/09a44b8cf0f9397465614e622979107f/raw/24-bit-color.sh >24-bit-color.sh

libdispatch efficiency tips

The libdispatch is one of the most misused API due to the way it was presented to us when it was introduced and for many years after that, and due to the confusing documentation and API. This page is a compilation of important things to know if you're going to use this library. Many references are available at the end of this document pointing to comments from Apple's very own libdispatch maintainer (Pierre Habouzit).

My take-aways are:

You should create very few, long-lived, well-defined queues. These queues should be seen as execution contexts in your program (gui, background work, ...) that benefit from executing in parallel. An important thing to note is that if these queues are all active at once, you will get as many threads running. In most apps, you probably do not need to create more than 3 or 4 queues.
Go serial first, and as you find performance bottle necks, measure why, and if concurrency helps, apply with care, always validating under system pressure. Reuse

Falsehoods programmers believe about time

This is a compiled list of falsehoods programmers tend to believe about working with time.

Don't re-invent a date time library yourself. If you think you understand everything about time, you're probably doing it wrong.

Falsehoods

There are always 24 hours in a day.
February is always 28 days long.
Any 24-hour period will always begin and end in the same day (or week, or month).

The Hugo static site generator sadly still uses the obsolete and badly standardized RSS format.

Here's how to set it up to generate an Atom feed instead. Pretty much all feed readers which understand RSS also understand Atom, except iTunes, and Atom is a better format.

Define an appropriate media type and corresponding output format in config.toml:

	"""
	The most atomic way to train and run inference for a GPT in pure, dependency-free Python.
	This file is the complete algorithm.
	Everything else is just efficiency.

	@karpathy
	"""

	import os # os.path.exists
	import math # math.log, math.exp

	private class CustomListCell: ItemListCell {

	private func defaultListContentConfiguration() -> UIListContentConfiguration { return .subtitleCell() }
	private lazy var listContentView = UIListContentView(configuration: defaultListContentConfiguration())

	private let categoryIconView = UIImageView()
	private let categoryLabel = UILabel()
	private var customViewConstraints: (categoryLabelLeading: NSLayoutConstraint,
	categoryLabelTrailing: NSLayoutConstraint,
	categoryIconTrailing: NSLayoutConstraint)?

	#include <stdio.h>

	// Compile with cc -o gray gray.c -framework UniversalAccess -F /System/Library/PrivateFrameworks

	extern void UAGrayscaleSetEnabled(int isEnabled);
	extern int UAGrayscaleIsEnabled();

	int main() {
	UAGrayscaleSetEnabled(!UAGrayscaleIsEnabled());
	}

	import UIKit

	@UIApplicationMain
	class AppDelegate: UIResponder, UIApplicationDelegate {

	var window: UIWindow?

	func application(_ application: UIApplication, didFinishLaunchingWithOptions launchOptions: [UIApplicationLaunchOptionsKey: Any]?) -> Bool {

	let window = UIWindow(frame: UIScreen.main.bounds)