slides

Getting Started

Use a template from either git submodule or direct clone method.

# git submodule add git@github.com:yhatt/marp-cli-example.git
git submodule update --init --recursive

rsync --archive \
    --exclude='.git' \
    --exclude='.github' \
    --exclude='assets/netlify-deploy-button.svg' \
    --exclude='LICENSE' \
    --exclude='netlify.toml' \
    --exclude='README.md' \
    marp-cli-example/ new-slides/

# git clone git@github.com:yhatt/marp-cli-example.git
# rm --recursive --force marp-cli-example/.git/

Local Development

Prerequisites

• Git
• Node.js LTS
• Python 3 (for the landing page dev server)

Landing page

The landing page (index.html) embeds a Google Maps map loaded from talks.toml. A dev server injects your API key at startup.

Set the key in whichever env file you prefer — dev.sh checks .envrc first, then .env:

# direnv (.envrc)
echo 'export GOOGLE_MAPS_API_KEY=your_key' >> .envrc

# or plain .env
cp .env.example .env
# Edit .env and set your Google Maps API key

If your .envrc contains secrets, add it to .gitignore.

./dev.sh
# Serving at http://localhost:8080

Override the port if needed:

PORT=3000 ./dev.sh

Or pass the key inline without any env file:

GOOGLE_MAPS_API_KEY=your_key ./dev.sh

Seasonal landing page

The seasonal landing page source is in seasonal/. The deployment publishes its build at the site root. It keeps the landing page content and layout and selects a background scene at build time from the seven day Oslo forecast supplied by MET Norway. It applies MET Norway's temperature thresholds for meteorological seasons. Each page refresh requests current Oslo weather after the initial content renders. The page shows a loading message until the live request succeeds and shows an unavailable message if it fails. Append ?season=spring, ?season=summer, ?season=autumn, or ?season=winter to inspect a specific scene.

npm --prefix seasonal install
npm --prefix seasonal run dev

The landing page uses the MET Norway Locationforecast Compact endpoint:

https://api.met.no/weatherapi/locationforecast/2.0/compact?lat=59.9139&lon=10.7522

Use Nushell to inspect the raw weather measurements:

http get 'https://api.met.no/weatherapi/locationforecast/2.0/compact?lat=59.9139&lon=10.7522'
| get properties.timeseries
| first
| get data.instant.details

For Nushell debugging, inspect the update time and all forecast periods:

let u = "https://api.met.no/weatherapi/locationforecast/2.0/compact?lat=59.9139&lon=10.7522"
let d = (http get $u)

let ts0 = $d.properties.timeseries.0
{
  updated_at: $d.properties.meta.updated_at
  instant: $ts0.data.instant.details
  next_1_hours: $ts0.data.next_1_hours
  next_6_hours: $ts0.data.next_6_hours
  next_12_hours: $ts0.data.next_12_hours
}

Individual slides

Each slide project has its own dev server powered by Marp CLI.

cd slides/demystifying-the-nix-store
npm install
npm start
# Opens a preview in the browser with live reload

Replace demystifying-the-nix-store with any of the other slide directories:

• governing-azure-resources-with-policy
• reproducible-environments-docker-vs-nix
• running-kernel-based-virtual-machine
• tracking-nixpkgs-merged-pull-requests

Attribution

Noto Color Emoji by Google is licensed under the Creative Commons Attribution 4.0 International License.

The seasonal landing page uses the Sylva Living World scene from Three UI.

See the Noto Emoji Animation documentation for technical details on animated emoji.