clubmirror v0.1.2: a static website generator for Strava club events

My local bike shop posts its group rides in a Strava club, but you need a Strava account to see them. So I wrote a small tool that publishes those events as a static page that they could link. It sat in a private repository for far too long because I kept thinking “it’s not ready yet”. But it’s public now, as clubmirror.

Actually it’s part of a process where I try to publish projects that were sitting in private repositories for far too long because I thought that they were not ready yet. This is the first one, but you can expect more of them to come.

What it does

clubmirror is a Rust command-line tool meant to run from cron/periodic. Each run it:

  • fetches the club’s group events from the Strava API,
  • renders them with Tera templates into plain static files,
  • writes an .ics calendar people can subscribe to,
  • downloads the GPX/TCX files of the routes attached to the events.

Then it exits, and that’s it. So there is no daemon, no database, no server-side code. It handles any number of clubs, each with its own templates, timezone and time window.

Try it without a Strava account

The repository includes a sample of fake events to play with:

git clone https://github.com/gawen947/clubmirror
cd clubmirror
chmod 600 examples/config.toml
cargo run -- -c examples/config.toml -v
python3 -m http.server -d public 8000   # open http://localhost:8000

Edit templates/default/, run again, and see the result. For a real club you need a Strava API application and a refresh token (cf README).

clubmirror is unofficial and not affiliated with Strava. Strava’s API terms restrict how its data can be displayed, so check them for your use.

What’s next?

  • Scripted per-club quirks with Rhai, instead of hardcoded ones
  • Proper debug options
  • Event filtering, for example keeping members-only events out of the calendar
  • Configurable event duration in the calendar
  • Processing clubs in parallel

runrunrun v0.3.0: I would use that

Another week, another release, runrunrun v0.3.0 is here. This release adds all the features I actually needed to start using this project myself on a daily basis.

What’s New

Fallback: The new -f/--fallback option (or RRR_FALLBACK=true) enables automatic fallback to previous matching rules when commands fail. This is useful when some commands aren’t found on the system:

https://* lynx
https://* chromium
https://* firefox

With fallback enabled, rrr -f https://example.com will try firefox first, then chromium, then lynx until one succeeds.

Alias redefinition: Redefining an alias now updates all rules using it, including previous ones:

[video] vlc
*.mkv [video]
*.mp4 [video]

# Now .mkv and .mp4 both use mplayer
[video] mplayer
*.avi [video]

Fix in desktop file support: Added support for more format specifiers (%f, %F, %u, %U) in desktop files (previously we only supported %U).

As always, feedback and contributions are welcome!


runrunrun is available at https://github.com/gawen947/runrunrun

runrunrun v0.2.0: desktop file import and more

Just one week after the initial release, here comes runrunrun v0.2.0 with a key feature from the roadmap: desktop file import.

What’s New

Desktop File Import: The :import directive can now read .desktop files and automatically generate rules from their MIME types:

:import /usr/local/share/applications/gimp.desktop
:import /usr/share/applications

This extracts the Exec and MimeTypes attributes, infers file extensions, and creates the appropriate glob patterns automatically. It’s a great bridge for migrating from existing desktop configurations.

Path Expansion: Configuration files now support tilde (~) and environment variable expansion:

:include ~/.config/rrr/work.conf
:include ~/$DOTFILES/rrr/common.conf

Include Loop Protection: Fixed a bug where circular includes would cause infinite loops. Now if a file has already been included, it’s simply skipped.

Performance: The config parser now only loads rules for the requested profile, improving startup time for large configurations.

Getting Started with Import

If you want to import your existing desktop file associations:

# Import a specific application
:import /usr/share/applications/firefox.desktop

# Import everything
:import /usr/share/applications
:import ~/.local/share/applications

Desktop files without Exec or MimeTypes are silently skipped, so you can safely import entire directories.

Next Steps

With desktop file import now available, migrating from traditional file association systems is much easier. As always, feedback and contributions are welcome!


runrunrun is available at https://github.com/gawen947/runrunrun

runrunrun v0.1.0 (first release)

I’m happy to announce the first release of runrunrun (v0.1.0), a new file and URL opener that runs the right thing—no surprises, no guesswork.

The Problem It Solves

If you’ve ever clicked on a text file only to have it open in Wine’s Notepad, or watched your carefully configured file associations get overridden by yet another desktop environment, you’ll understand the frustration. Every time I need to configure file associations with xdg-open and its cousins, something isn’t quite right.

Traditional desktop openers work by making assumptions—and that might be exactly what some users want. But for those who prefer explicit control, the current tools fall short. They guess your preferred browser based on your desktop environment, scatter configuration across countless .desktop files, and make debugging file associations unnecessarily complex.

The goal of runrunrun is simple: make opening files and URL schemes straightforward and consistent.

What Makes runrunrun Different

runrunrun (or rrr for short) takes a radically simple approach: explicit configuration through pattern matching. No MIME type cascades, no desktop file archaeology—just straightforward rules that you control.

Here’s what opening a PDF looks like in your config:

*.pdf    qpdf

That’s it. When you run rrr document.pdf, it opens in qpdf. No surprises.

Core Features in v0.1.0

Simple Pattern Matching: Define what opens what using glob patterns for files and URLs:

*.jpg        feh
https://*    firefox
mailto:*     thunderbird

Predictable Overrides: Later rules win, making customization straightforward:

*.txt    mousepad
*.txt    vim    # This one wins

Regex Support with Capture Groups: For more advanced matching, use regex patterns (prefixed with ~) that can even extract and reuse parts of the match:

~^IMG_[0-9]+\.png$  darktable
~mailto:([^?]+)\?subject=([^&]+)  "true %s; thunderbird --compose \"to=%1,subject=%2\""

Profiles for Different Contexts: Switch between configurations easily:

:profile work
https://*    firefox-work-profile

:profile personal
https://*    brave

Terminal-Friendly: Unlike desktop-centric tools, rrr works just as well in terminal environments as it does on a full desktop, making it suitable for servers and minimal setups.

Transparent Operation: Query mode (rrr -q file.ext) shows you exactly what would run, and dry-run mode (rrr -n) lets you test configurations safely.

Getting Started

The project is written in Rust and available on GitHub. Configuration lives in simple text files at /etc/rrr.conf or ~/.config/rrr.conf.

What’s Next

This first release includes the core functionality along with regex patterns and aliases for common applications. Future versions will add the ability to import existing .desktop files for those who want a migration path from their current setup.

As with any first release, there are likely bugs to catch and behaviors that might need adjustment. Your feedback will help shape the direction of future releases.

If you’re tired of fighting with file associations and want something that just works the way you configure it, give runrunrun a try. Your feedback and contributions are welcome!


runrunrun is available at https://github.com/gawen947/runrunrun