in reply to Music shuffling

I think it's a fair approach. You are making sure that any of N songs isn't played more often than every N/2 songs. Though depending on the size of your library you could as well play them strictly round robin and wouldn't notice any difference.

One variation of your method might be to use something like(rand()**3)*($N*.95). This makes it possible but highly unlikely for a song just played to play again quite soon. rand()**3 strongly favors songs at the beginning of the list

Personally I use a jukebox program where I put the songs in directories that were given different weights. If there were directories a, b and c, the files in b would be played 2 times as often as the songs in dir c, and the songs in a 3 times as often. The ordering is simply alphabetical

Replies are listed 'Best First'.
Re^2: Music shuffling
by oko1 (Deacon) on Jul 08, 2008 at 16:17 UTC
    you could as well play them strictly round robin and wouldn't notice any difference.

    Well, yes - but even with my (fairly large) collection, I'd still have a problem with hearing a pattern after a while. This approach can run forever without that. I suppose that I could also play the entire list, shuffle it, and then play it again, with a small possibility of a repeat at the "join" - maybe even reshuffle if any of the X last-played songs are found in the first X entries of the new list - but I didn't think of that approach until just now. :) 

    
-- 
Human history becomes more and more a race between education and catastrophe. -- HG Wells
[reply]

In Section Seekers of Perl Wisdom