Introduction

potcasse, pronounced "pot kas" is meant to help people to publish and self host a podcast easily without using a third party service.

Dependencies

• rsync (could use cp but avoid recopying audio files locally).
• a posix compatible OS (Linux, *BSD, Solaris).
• some webserver and/or some gemini server to host the files.

How to use

The idea is to regroup audio files with their metadata in a directory and generate the structure that you will publish on a web server and/or on a Gemini server.

A simple index.html file and a simple index.gmi file are also generated in the process to give an easy list without using the RSS file.

First time

potcasse init

This will create a file metadata.sh that you need to edit accordingly to your setup.

It has the following variables:

• TITLE: this is the podcast title.
• AUTHOR: this is the podcast author (doesn't support multiples authors yet).
• SITE: base HTTP URL where your podcast will be available (for example https://tilde.example/myname/podcast/.
• CAPSULE: base Gemini URL where your podcast will be available (for example gemini://tilde.example/myname/podcast/.
• DESCRIPTION: description of the podcast.
• RSSLINK: name of the RSS feed.
• IMAGE: if value is not empty, potcasse will use the file logo.png (only used on the web version)
• LANGUAGE: language code (such as fr or en-us) that can be potentially used by some players.

You will share the link $SITE/index.html, $SITE/index.gmi or $SITE/$RSSLINK to your listeners.

New episode

potcasse episode "Episode XX: trying something weird" /path/to/audio/file

This will create a directory episodes/YYYYMMDDhh so you can't publish more than a podcast every hour using this command. The audio file inherits the directory name.

You can force a directory name with a parameter after the audio file.

potcasse episode "Episode XX: trying something weird" /path/to/audio/file this_is_my_podcast_episode_i_want_to_name

Publishing

potcasse gen

This will create or update the output_html directory with your audio files, the RSS file, an index.html file listing all the episodes and the logo file if any. It will also create or update the output_gmi directory according to the same scheme, in order to create a Gemini compatible version of the listing (it will not use the logo file though).

Real world example (from solene)

My podcast feed is available at https://perso.pw/lambda/feed.xml which is on server perso.pw in /var/www/htdocs/lambda/.

I add new episodes with potcasse episode "A made something nice" ~/Podcast/episodeXX/file.ogg then I use potcasse gen that will refresh the directory public_html with metadata and episodes, finally I put it online with rsync -av public_html/ perso.pw:/var/www/htdocs/lamda/.

My metadata.sh looks like this:

TITLE="Lambda Solene"
AUTHOR=Solene
SITE=https://perso.pw/lambda/
RSSLINK=feed.xml
IMAGE=YES