dashboard

Licensed under a Creative Commons Attribution 4.0 International License.

---

A lightweight, self-refreshing information dashboard designed for a fixed display in a living room, running on a Raspberry Pi. No server, no framework — just a Python data fetcher and a static HTML page.

Features

• Current weather — temperature, wind speed, precipitation probability, sunrise/sunset times, weather icon (WMO code)
• 6-day forecast — daily min/max temperatures, weather icon, rain probability and accumulation
• Air quality — real-time European AQI with colour-coded label (green → amber → red)
• News feed — 2×2 grid: 2 local headlines (Ouest-France, Rennes) and 2 world headlines (Courrier International), full titles, no truncation
• Live clock — refreshes every 30 seconds without reloading the page

File structure

dashboard/
├── data/
│   └── dashboard.json     # generated by update.py, read by app.js
├── venv/                  # Python virtual environment (created by install.sh)
├── .firefox-profile/      # Firefox profile for local file access (created by start.sh)
├── app.js                 # Renders dashboard.json into the three panels
├── index.html             # Page shell; loads fonts, Weather Icons, app.js
├── style.css              # Dark theme, CSS tokens, responsive layout
├── update.py              # Fetches weather + news, writes dashboard.json
├── install.sh             # One-time setup (venv + dependencies)
├── start.sh               # Main launcher (data fetch + browser + refresh loop)
├── dashboard.desktop      # Double-click entry point for the file manager
└── start.log              # Runtime log written by start.sh (auto-created)

How it works

update.py calls two free APIs (no key required):

• Open-Meteo — weather forecast and current conditions
• Open-Meteo Air Quality — European AQI

It also parses two RSS feeds:

• rennes.maville.com — local Rennes news (Ouest-France)
• courrierinternational.com — world news

All data is written to data/dashboard.json. app.js fetches that file on page load and renders the three panels. The page itself never makes network requests — all external calls go through update.py.

start.sh orchestrates the full lifecycle: it runs update.py once on startup, opens the browser in kiosk mode, then re-runs update.py every 15 minutes in a loop and reloads the page (via xdotool).

Requirements

• Raspberry Pi running Raspberry Pi OS (or any Debian-based Linux)
• Python 3.7+
• Firefox or Chromium
• Internet connection (for APIs, RSS feeds, and CDN-hosted fonts and icons)

Optional but recommended for automatic page reload:

sudo apt install xdotool

Installation

Run once, from inside the project directory:

cd ~/dashboard
bash install.sh

This creates the venv/ folder, installs requests and feedparser, and generates the first dashboard.json.

Usage

Edit dashboard.desktop and replace the path with the actual location of your project:

Exec=bash /home/your-username/dashboard/start.sh

Then make it executable (right-click → Allow executing in the file manager, or):

chmod +x dashboard.desktop

Double-click dashboard.desktop to launch. The dashboard will open full-screen and refresh automatically every 15 minutes.

All runtime output is logged to start.log in the project directory — useful for troubleshooting.

Configuration

To use a different location, edit the coordinates at the top of update.py:

LAT      = 48.1173        # latitude
LON      = -1.6778        # longitude
TIMEZONE = "Europe/Paris" # IANA timezone name

To change the refresh interval, edit INTERVAL in start.sh (value in seconds):

INTERVAL=900  # 15 minutes

To add or replace news sources, edit the news dict in update.py and update the matching key lookups in app.js:

news = {
    "Rennes (Ouest France)":          rss_titles("https://..."),
    "Monde (Courrier international)": rss_titles("https://..."),
}

Design

• Fonts (Google Fonts CDN): Bebas Neue (clock, large temperatures), Inter (body), Roboto Mono (numeric values in forecast)
• Icons: Weather Icons 2.0.10 via cdnjs — mapped from WMO 4677 weather codes
• Colour palette: near-black background (#0d0f14), orange accent for warm values (#f97316), blue for cold/rain (#60a5fa), semantic colours for AQI
• Layout: CSS Grid, two columns on top + full-width strip at the bottom; responsive single-column below 900px