i'm learning quite a bit from audiomatch. it becoming relatively popular, i've had to resort to peddling for donations ... and also taking a hard look at my code to see how to optimize stuff.

i wrote in a bunch of benchmarking in the script that handles the data output to track where all the time goes. it was apalling to me to see that some queries were tkaing up to 30 seconds to add a song.

and now i know why people complain. but in any case, i've taken a short look at the data, and it seems rather sporadic; sometimes the image generation takes a bulk of the time, and other times, it's just database work.

surprisingly mysql handles indexed fields (text fields ... of about 150K entries) pretty darn well.

but in all honesty, i don't see how the script will continue to handles so much data in the upcoming weeks. i'm going to do my best tomorrow to identify the bottlenecks within the script and to relieve congestion. i already located a huge bottleneck ... time to go byebye ;)