Brand new podcast platform, Spooler. Uses podcast:guid and podcast:locked - doesn't use JSON chapters, but does use in-MP3 ones. Quite technically interesting. https://www.spooler.fm/press-spooler-launch.html
@jamescridland Tracked by Chartable (a Spotify company) for analytics.
Audio delivered in 128kbps 48khz mono
ffmpeg used
Embedded id3 chapters for the short segments (text only no images)
Huge embedded id3 apic image for the podcast show art up front (before any audio frames): 1.1mb 3000x3000 jpeg, out of the 10.7 mb mp3 file
Not sure how dynamic it is (per request), I think they just rebuild the static audio often ("editions").
Served by cloudflare, origin on S3? has the x-amz- headers
@jamescridland At the end of the first day, the podcast produced only 5 "editions" of the short (around 10 minute) podcast episode.
4 preliminary editions, and one final. BBC has feeds that update far more often than that.
Interesting that the preliminary episode links are now dead, for example edition 3 [1]
