One problem that really interests me is the "dynamic playlist." Once your music library becomes sufficiently large and diverse, simply shuffling songs ceases to be an acceptable way to listen to music. You'll randomly switch between vastly different genres, come across tracks that aren't enjoyable outside the context of an album, jump in halfway through some pieces. It's a mess.
I wrote a simple script which uses Last.fm data to generate a "path" through the artists in my music library based on their similarities. It's very far from perfect, but it suffices to build an album playlist which slowly takes me through several genres. Some day I'll improve it to work based on albums rather than artists.
Ideally, I'd like to be able to run queries on my music library and have an interesting playlist returned to me.
> Ideally, I'd like to be able to run queries on my music library and have an interesting playlist returned to me.
You want Beets.[1] Anything you could possibly want is doable with the smart playlist feature[2] and the redonkulous power of the queries and metadata (chromaprint/musicbrainz/echonest). There is a plugin for last.fm genres but I have not used it. I am not sure how you got similarities from last.fm but you can definitely use the echonest data to find similar music. There was a good discussion of beets here not so long ago.
I wrote the play plugin for beets last weekend that turns a query into a temporary playlist and opens it with your systems music player for just this reason.
I've been meaning to work more on this, but my free time is fairly limited. At the very least I should probably put my script on GitHub (although it's far from groundbreaking).
For me, one of the pain points is quality of metadata. It's all very well collecting genres, tags, and similar artists from a service like Last.fm, but you need to be reasonably confident that they align with your own ideas about music. Since everyone's ideas about music are different, this puts you in a tight spot.
rateyourmusic.com's genre system works pretty well, it's based on user voting on genres for an album (genres are selected from a list, which is constantly being updated by democratic process too)
I wrote a simple script which uses Last.fm data to generate a "path" through the artists in my music library based on their similarities. It's very far from perfect, but it suffices to build an album playlist which slowly takes me through several genres. Some day I'll improve it to work based on albums rather than artists.
Ideally, I'd like to be able to run queries on my music library and have an interesting playlist returned to me.