The SQL collection loads first and sets the metadata of the existing tracks in the playlist.
This should not happen. The metadata of those existing tracks should be fetched from the Nepomuk collection or the fetching part should wait until the Nepomuk Collection fully loads.

In the edit track details, the bitrate of a track is shown in bps rather than kbps that most people use.

Make it kbps.

The current Amarok::Component::logger doesn't work with ThreadWeaver Jobs.

Bart said,

Only KJobs (include KIO::Job) and cutom QObjects which emit the right signals. You can use the last one to make it work. If there are multiple threadweaver jobs that need to do the same (perhaps multiple inside your code) you should add support for it to Logger.

So implement either one of these solutions and get the Progress bar working.

The SQL collection shouldn't even show up in the collections pane.

Only the Nepomuk Collection should be visible.

Line 368, NepomukMetaConstructJob

Instead of defining a new mapchanger object for every track, move it outside the loop so that one mapchanger object is used to add all tracks.

Create an ontology for score in Nepomuk and then start handling scores as you handle ratings.

Port to Nepomuk2 and

1. Get rid of all the warnings.
2. Start using ResourceWatcher

Nepomuk collection displays tracks from all over the system and does not stick to the folders that Amarok is supposed to index/watch.

This is a major feature that MUST be fixed.

Use Q_UNUSED for variable newScore until scores are handled.