monline

Some Python scripts use the webbrowser module to launch a browser for a given URL. If you're on a headless server, it'll launch a text-based browser (eg. in Debian it's w3m) which might not be desired (many websites are less than usable in text-based browsers and the launch of such an URL in a text-based browser is more an annoyance than help). You can easily avoid this behaviour.

The documentation of the webbrowser module says that one can override the module's browser detection/selection algorithm by setting the BROWSER environment variable. Thus the solution is fairly trivial:

Eg. scripts for various online services (Google+, Facebook, Flickr, etc.) need authorization and try to open an URL. These scripts print the URL to the terminal and open the URL via webbrowser at the same time. Since you can easily copy&paste the URL from your terminal into a browser, the text-based browser part is unnecessary.