Repository navigation
Description
We want to set up a numbers station - we can probably stream this via nginx.
The station should serve some audio 24/7, but allow us to intersperse messages
into the output every once in a while. It should be possible to automatically
generate messages, e.g. via espeak or fixed recordings of DevOps team members
reading out numbers and letters.
Since Joe has radio experience, he can probably help a large part in
implementing this.
Goals
-
Continuously endpoint most likely via nginx
-
Command to intersperse output (mp3 for a start)
-
Command to automatically speak output (can be espeak for a start)
Scheduling can be performed via cron or at.
Components
-
Icecast to host the audio over HTTP
-
VLC loops audio. Should expose the telnet interface for pausing and resuming.
-
ALSA with virtual sound devices
-
Festival for speech synthesis
-
Timer to regularly play propaganda
-
ffmpeg
-
Plays propaganda to the stream
-
Listens to VLC and outputs to Icecast
-
Metadata
Metadata
Assignees
Labels
Type
Projects
- StatusShow more project fieldsIn Progress
Activity
jb3 commentedon Aug 20, 2024
Python Discord, it's time.
jb3 commentedon Aug 20, 2024
Suggested: We have decided to infringe on Lemon's copyright (with attribution). We will use Beardfist for the static music playout.
jb3 commentedon Sep 1, 2024
For reference: https://discordier.github.io/sam/
jchristgit commentedon Sep 7, 2024
I am unassigning myself, as implementing the mail authentication daemon will take the majority of my DevOps Sprint.
Xithrius commentedon Apr 15, 2025
how many story points @jb3
Xithrius commentedon Apr 15, 2025
crank.that.soulja.boy.spongebob.webm
jchristgit commentedon Feb 27, 2026
Implemented and broadcasting propaganda.
jchristgit commentedon Feb 27, 2026
Missing Ansible setup.
jchristgit commentedon Mar 26, 2026
As discussed, this will be implemented on turing to trial out Guix.