Skip to content

QUICKSTART.md

Latest commit

 

History

History
82 lines (60 loc) · 3.02 KB

QUICKSTART.md

File metadata and controls

82 lines (60 loc) · 3.02 KB
Raw

Quick Start

Extended version of the README quick start, for a first run.

Requirements

  • OS: Windows 10 or 11 (primary). The GUI needs Edge or Chrome — both ship with Windows. macOS/Linux run the engine and the CLI fine, start.bat is Windows-only.
  • Python: 3.10 or newer
  • Disk: ~150 MB for the Playwright Chromium — only ever used as the datanodes fallback when no real Chrome is found. fuckingfast needs none of it.
  • Network: any speed. The defaults assume fiber; lower DL streams on a slower link.

Option 1 — One-click (Windows)

  1. Install Python 3.10+ with "Add Python to PATH" ticked.
  2. Download or clone this repo.
  3. Double-click start.bat.

First run installs aiohttp, playwright and curl_cffi, then the Chromium build. start.bat then starts a loopback HTTP server and opens the GUI in Edge (or Chrome) with --app: a window with no tabs and no address bar. There is no native GUI dependency.

Option 2 — Manual

git clone https://github.com/LeyckerS/moondownloader.git
cd moondownloader

pip install -r requirements.txt
playwright install chromium          # datanodes fallback only

python moon_bridge.py                # GUI
python moon_bridge.py --serve        # server only, prints the URL

Option 3 — CLI (headless)

python moon_cli.py --urls links.txt --output ./downloads --browsers 16 --streams 48

--browsers is the number of parallel extraction workers, not browsers: Chrome opens once and only if a datanodes link shows up. python moon_cli.py --help for the rest.

Just want to look at the interface?

Open web/index.html in Chrome or Edge. It boots in demo mode against a synthetic engine — no Python, no downloads.

First run

  1. Paste datanodes.to and/or fuckingfast.co links into the editor. The count and the per-host split update as you type.
  2. Pick an output folder (defaults to ~/Downloads/datanodes).
  3. Leave the settings alone for a first run. If you want to tune:
    • Extractors (2–32) — parallel extraction workers
    • DL streams (2–48) — concurrent downloads; fewer means more bandwidth each
    • Pages (1–8) — datanodes only: tabs on the one shared Chrome window
    • Captcha wait — datanodes only: how long a manual Turnstile solve may take
  4. Download, or switch to Links only to extract direct URLs without downloading.

A fuckingfast-only run never opens a browser window. The first datanodes link opens one Chrome; solve the Turnstile there if it asks, and the clearance is reused for the rest of the session.

What a run leaves behind

Next to the scripts:

File Contents
moontech_*.log human-readable performance report
moontech_*.json per-file metrics
output_links.txt extracted direct links (Links-only mode)
failed_links.txt URLs that failed every retry
settings.json GUI settings, pasted links, language

Interrupted files stay as .tmp and resume on the next run via a Range request.