PodOsef is an MP3-first podcast publishing system: drop MP3s into a folder, it generates HTML + RSS. No DB, the MP3 metadata is the source of truth. Also supports an archive mode that watches a main feed and mirrors episodes. Repo + write-up:
https://meirz.net/2026/02/22/podosef-mp3-first-lightweight-simple-podcast-publishing-system/
I deleted previous post as I managed to screw it up pretty good. There was a comment that the entire “mp3-first” idea is outdated… That might be true. However I developed it based on my requirements. I think if there is enough interest, software/design can be updated.
By the time I finished with the software, I had at least few different ideas for the future, but I stopped where I stopped, just because I got everything I needed. I keep my mind open, if there is enough interest, perhaps software can be enhanced and developed further.
The banner image looks like AI.
Do you mean mean image in the article? Yes, generated by AI. I needed image for the project and a sample podcast cover (apple spec). Unfortunately I would not be able to draw anything except for simple smiley even if my life dependent on it… I’m not gifted in drawing/graphics are at all.
Is there anything wrong with it?
Here are a ton of royalty free podcast themed images: https://unsplash.com/s/photos/Podcast
Yeah, it’s ugly and the geometry makes no sense. How have we as a species already forgotten about stock photography?
For the post I simply reused sample rss cover but scaled it down to 1024 x 1024 pixels. Original is 1400 x 1400 pixels and 483 KB - by Apple spec must be 1400 x 1400 to 3000 x 3000 pixels and I opted to keep size under 500 KB (to avoid any issues). I know it is on smaller size, but it is just a sample cover - should be replaced during real deployment, but otherwise it should work properly as a demo.
Thank you.
After feedback on artwork, one of my friends gave me new artwork for the podOsef. I updated the post and source code.