The iPod server was a powerful first iteration music sharing device. While not designed for convenience, its internal RAID can still be expanded with modern compact flash cards for 1.25TB of on-the-go storage.
Image transcription: A photo of an early iPod with colour display running macosx server 1.2. It looks cumbersomely thick, but also desirable.
(Originally published earlier today on bitbang.social)
The track data and everything is stored in the file, the file name is irrelevant.
Iirc thats what OP is referring to, they would strew the files stripped of meta data across the partition in various folders with little discernible structure, track data was written to the database and would be queried on song plays. I believe the advantages were faster bootup (Scan db instead of thousands of files) and the random file names were to ensure unique keys in db and help the “random” algorithm.
iPods didn’t even strip the files of metadata, it just renamed them on disk