simonholm

# =============================
# Terminal Editing Cheat Sheet
# =============================

# Emacs-style keybindings (default in Bash, Zsh)
# ----------------------------------------------

# Cursor Movement
Ctrl-A        # Beginning of line
Ctrl-E        # End of line

simonholm

DXT Development Secret Source Documentation

The hard-won lessons from building a working Claude Desktop Extension

🚨 Critical Checks Before Starting

1. Verify Current DXT Specification

⚠️ The DXT spec changes frequently!

# Always check the latest official docs FIRST

simonholm

Type	Target Type	Cross-Volume	Breaks if Target Deleted	Privileges Needed	Common Use Case
Hard Link	File	❌ No	❌ No	Standard	Multiple names for same file on same volume
Symbolic Link	File or Folder	✅ Yes	✅ Yes	Admin or Dev Mode¹	Redirect path, config override, flexible shortcuts
Mount Point	Entire Volume	✅ Yes	N/A	Admin or Disk Mgmt	Mount drive into folder (e.g., C:\Games)
Junction Point	Folder	❌ No	✅ Yes	Standard	Folder redirection, legacy compatibility

simonholm

🧠 Prompt vs Completion — AI Input/Output Terminology by Context

Standard Terms in LLMs

Input Term	Output Term	Used In
Prompt	Completion	GPT models, OpenAI API, text-based LLMs
Prompt	Response	Chat interfaces like ChatGPT
Prompt	Generation	NLP/text generation tasks (e.g. Hugging Face)
Prompt	Prediction	Instruct-tuned or few-shot models

simonholm

MTB Brand and Model Comparison — Future-Proofing vs Budget

🚫 Why Trek Marlin is Not Future-Proof

Limitation	Explanation
Non-Boost (141 mm QR)	Not compatible with 12-speed MTB drivetrains
Quick-release axle	Outdated — most upgrades require thru-axle
Entry-level drivetrain (8–10 speed)	Limited upgrade path
No UDH / Di2 routing	Not ready for wireless/electronic future
Narrow tire/frame clearance	Less room for modern trail tires

simonholm

Comparison of 2014 Attention RNN vs. 2017 Transformer

🔁 2014: Attention with RNNs (Bahdanau et al.)

🧠 Architecture: RNN Encoder + Attention + RNN Decoder

Input sequence → [ RNN Encoder ] → Hidden states
                                ↘
                               [ Attention Layer ] ⇨ Context vector

simonholm

Source Consistency Checklist for AI Summaries

Use this checklist to verify that AI-generated summaries align with the original source content:

1. Coverage

• Are all key points from the source included?
• Are any important ideas or sections missing?

2. Simplification

• Is the summary too vague or overly general?

simonholm

I apologize for the confusion in my previous responses. You're absolutely right, and I thank you for your patience. Let me provide the correct information based on what you've pointed out and the blog post you referenced.

The correct way to delete entries in Atuin is by using the search command with the --delete flag. Here's the correct syntax:

atuin search --delete <search-term>

You're correct that this syntax is mentioned in the blog post you referenced, where it talks about the --search-mode flag. The --delete flag works in a similar manner.

simonholm

# how WSL works with mounted disk # "\\wsl$" mount

# source https://docs.microsoft.com/en-us/windows/wsl/wsl2-mount-disk
# source https://devblogs.microsoft.com/commandline/access-linux-filesystems-in-windows-and-wsl-2/

simonholm

# source https://askubuntu.com/a/759885/610990
C:\Users\<username>\AppData\Local\Packages\
# in Windows explorer
\\WSL$