Everyone is welcome to edit the wiki, please help yourselves

• Logging – should go to a file or file+stdout
• headers field from resultsJSON should add headers to the http:request

• use new connection for each file transfer so we don’t have so many flow control issues when you want to dispatch queries but the connection is full of audio data.
• integrate upnp port-fwd setup as a driver (miniupnpc)

• integrate mxcl’s inotify/mac-equiv driver for insta-scanning. figure out if we can do it on windows too.
• store dirs to be monitored somewhere (config file/fixed dir of symlinks? needs to be configured by a gui)

• tidy up, build an internal api so other apps can use it, then make p2p bootstrap the mesh network by negotiating connections via xmpp msgs.

• Provide a “can not download” option for streaming sites that don’t allow saving per theirTOS (Rhapsody).
• Can a resolver provide a url to a “listen” page that’s not an MP3? Unfortunately looks like we can’t stream directly –RhapsodyTOS

• Hosted magnatune search index so clients can resolve free magnatune tracks
• Spotify resolver (use despotify (premium subcribers only). or viaspotify metadata api? )
• Rhapsody resolver (USA only?)
• Sexier web + desktop demos
• Unit testing?
• DAAP resolver (ie Firefly Media Server, etc)

XSPF urls can be handled by playlick or playgrub, like so: http://www.playlick.com/#xspf=www.example.com/playlist.xspf

Create xspf_opener.sh in your /usr/bin:

#!/bin/bash
firefox “http://www.playlick.com/#xspf=$1”

Go to firefox preferences, applications, and change the default app forXSPF to this script.
If not set, click anXSPF in firefox and set it to remember to use xspf_opener.sh

ExampleXSPF file