0.14.0's database and external apps #629
-
|
I've started a little personal project that is essentially a listen database with LB's api, and activity pub support. But now that multiscrobbler has its own database, I kinda want to directly use multiscrobbler's database instead of reinventing the wheel. But for that, I got a few questions regarding it:
That would be nice to know before I go too deep into it |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
|
Hi! The database is not really intended for external access at this time. The current design is meant for ephemeral play/scrobble only because I (currently) don't want MS being a full-fat scrobble server/storage location. Due to this there are many design decisions that are not beneficial for long-term storage such as tracks/artists/albums not being normalized into separate tables. As for access to play/scrobble data for a "recent" window (within whatever you have set for the retention policy), I am in the middle of a redesign of the ui that will introduce more stable/powerful api endpoints that I plan on documenting. With those in mind:
Yes, you will be able to access discovered plays, scrobbles, and now-playing data for each component. Using an API endpoint that returns JSON. With parameters for filtering by properties as well as a time range.
MS will have concepts for discarding listens using a transform stage in a future release. Additionally, the should scrobble behavior will be soon be implemented for all sources/clients so that you can "mute" a component for some period so it does not either discover listens, or record a scrobble. However I don't have plans for directly deleting listens/scrobbles. Or allowing that kind of behavior to propagate downstream (IE delete a scrobble in MS -> deletes the scrobble in Koito) The TL;DR MS's design, and the database, are still focused on being the "glue" between your listening sources and other, separate scrobble storage services. You will soon be able to access your recent (like 7 days, maybe a month) of listens via an API but MS is not a good choice for permanent scrobble data. |
Beta Was this translation helpful? Give feedback.
Hi! The database is not really intended for external access at this time. The current design is meant for ephemeral play/scrobble only because I (currently) don't want MS being a full-fat scrobble server/storage location. Due to this there are many design decisions that are not beneficial for long-term storage such as tracks/artists/albums not being normalized into separate tables.
As for access to play/scrobble data for a "recent" window (within whatever you have set for the retention policy), I am in the middle of a redesign of the ui that will introduce more stable/powerful api endpoints that I plan on documenting. With those in mind: