3 ms·
There's definitely a feeling that track orders feel quite similar between play attempts because there isn't a large enough "shuffle memory" between sessions, or
by WorldMaker 7y ago
There's definitely a feeling that track orders feel quite similar between play attempts because there isn't a large enough "shuffle memory" between sessions, or in many of these services any "shuffle memory" at all (some really are just random and hoping the distribution of the random function enough to compensate).
5 Disk Changers had a finite need for memory for recalling what had been played, which was easy enough. Some people's libraries are so large it's an interesting question of how much is it worth storing the played state of a song, and for how long (can you clear that to start a fresh shuffle? can you add individual songs back without just clearing the memory state?).
Some players let you manually manage this: shuffle songs into a shuffle playlist and then play that list in track order, remembering your last heard state.
(For the purposes of dealing with the bad PRNG in my car I wrote a Python script a few years ago to do a version of that shuffle playlist management in how it fills thumb drives intended for my car. It keeps a shuffled track list in YAML and then uses that copy files to the drive by track number for the simple lexicographic default play order of the car.)