At first, I wrote this introduction to this entry:
I know what kind of iPod Shuffle accessory would make it really, really great. I'm not sure what it would look like, how it would work, or how much it would cost, but I still want it: an auxiliary entropy source. Today, I found myself listening to the same handful of artists from the beginning of the alphabet. I'll start my Shuffle now and we'll see what it wants to play. I'll stop when something doesn't fall into the pattern:
By this point, though, I got suspicious. I stopped to review my actual playlist. My Shuffle, which is supposed to contain everything from my Shuffle playlist -- that's 103 songs at the moment -- instead contains thirty-five songs. Some of them, like Bjork, don't even appear in my Shuffle playlist at the moment.
Sometimes I really, really wish that the iPod devices kept their database of music in a really easy to read (documented) format. If they did, I could easily write my own software to build and sync playlists. I could still have playlists like those I've enjoyed for so long, but I could make them much more enjoyable with constraints like, "no more than n% of playlist contents may be from a single artist."
When a friend of mine started talking about getting an iAudio recently, I replied to his LVLUG posting saying that I was too addicted to the "last played" and "play count" and "rating" features supported by iTunes and the iPod players. This is both a blessing and a curse. I can build really useful smart playlists, but I'm chained to the slow, bloated iTunes player when working at my laptop, and I'm stuck with lousy syncing on the shuffle and glaring feature omissions on all the smart playlists. Don't get me started on the fact that iTunes still doesn't have a "Update Playlist Contents" context menu item for non-auto-updating smart playlists.
Will someone please release a reasonable mp3 player that keeps track of these things? I am longing to abandon iTunes, but until something else can deal with iPod's database, or until another player can produce this kind of data, I am too addicted.