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)
I once had a Mac guy turn up at my workplace and tell me ‘the document you need is on this iPod’.
I plugged it into a Linux machine and it mounted the filesystem just fine.
The horrors I saw in that filesystem, how it was laid out and how the iPod needed to use its database to present real names to the user …
I never want to see that again.
I mean in Apple’s defense you wouldn’t want a media player to show filenames to the user anyways, you would want to display the artist/track name from the mp3 tags. The 4 letter filenames are a hash table presumably due to length restrictions in the firmware and/or performance reasons
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