#metabrainz

/

      • reosarevok[m]
        As in: we want to do that, can anyone see any reason why we shouldn't
      • 2025-06-30 18134, 2025

      • reosarevok[m]
        But ideally that'd also include people missing today such as mayhem so maybe I'll ask again later :)
      • 2025-06-30 18147, 2025

      • reosarevok[m]
        If anyone is against, or has any comments, please say so in the ticket :)
      • 2025-06-30 18152, 2025

      • reosarevok[m]
        Anyway, I think it's good to stop for today!
      • 2025-06-30 18158, 2025

      • reosarevok[m]
        Thanks everyone for being around
      • 2025-06-30 18128, 2025

      • reosarevok[m]
        </BANG>
      • 2025-06-30 18116, 2025

      • nik[m] has quit
      • 2025-06-30 18131, 2025

      • suvid[m]
        kellnerd... (full message at <https://matrix.chatbrainz.org/_matrix/media/v3/download/chatbrainz.org/FaQxeNYPLWrEFzGRpZilVqCl>)
      • 2025-06-30 18135, 2025

      • suvid[m]
        lucifer: should i just call spotify web api for every listen as it gives all data for a track... (full message at <https://matrix.chatbrainz.org/_matrix/media/v3/download/chatbrainz.org/cYSDXokXFxcQrHFalHDambZa>)
      • 2025-06-30 18154, 2025

      • lucifer[m]
        suvid: we want to use the spotify metadata cache when possible to avoid making api calls. in production ideally api will never be called because the cache is supposed to be comprehensive.
      • 2025-06-30 18122, 2025

      • suvid[m]
        lucifer[m]: does the cache have duration_ms?
      • 2025-06-30 18131, 2025

      • lucifer[m]
        it does but you don't need that
      • 2025-06-30 18139, 2025

      • suvid[m]
        yes musicbrainz data is very comprehensive so idts we would need spotify web api calls
      • 2025-06-30 18141, 2025

      • lucifer[m]
        it will be in the data field.
      • 2025-06-30 18157, 2025

      • suvid[m]
        lucifer[m]: but it will be used in validating the listen no?
      • 2025-06-30 18108, 2025

      • lucifer[m]
        like i said we are not using musicbrainz data. the cache is totally separately from musicbrainz and is not used in it.
      • 2025-06-30 18123, 2025

      • suvid[m]
        lucifer[m]: ohh ok
      • 2025-06-30 18143, 2025

      • lucifer[m]
        the cache is not guaranteed to have all items so in that case we should have api calls as fallback.
      • 2025-06-30 18157, 2025

      • lucifer[m]
        it will ideally have 99.9% of the items but we have never tested it.
      • 2025-06-30 18158, 2025

      • suvid[m]
        👍️
      • 2025-06-30 18101, 2025

      • lucifer[m]
        in that context.
      • 2025-06-30 18109, 2025

      • suvid[m]
        okay i'll get duration from cache first
      • 2025-06-30 18114, 2025

      • lucifer[m]
        for validating the listen, for now consider all listens as valid.
      • 2025-06-30 18123, 2025

      • suvid[m]
        lucifer[m]: 👍️
      • 2025-06-30 18138, 2025

      • lucifer[m]
        for duration i need to check the docs on whether we want the duration actually played or the full duration of the song.
      • 2025-06-30 18152, 2025

      • lucifer[m]
        for now assume the duration in the dumps is enough but leave a todo in the code.
      • 2025-06-30 18117, 2025

      • suvid[m]
        left a WIP in the code for the validate listens function
      • 2025-06-30 18125, 2025

      • suvid[m]
        * a WIP comment in the
      • 2025-06-30 18148, 2025

      • suvid[m]
        <lucifer[m]> "query spotify_cache.track with..." <- added the sql query (should work hopefully but didn't test as of now)
      • 2025-06-30 18149, 2025

      • suvid[m]
        took a while
      • 2025-06-30 18149, 2025

      • suvid[m]
        also took help of gpt as it got very complicated 😢
      • 2025-06-30 18126, 2025

      • suvid[m]
        lucifer: Could you pls check the PR
      • 2025-06-30 18126, 2025

      • suvid[m]
        I have made some significant progress in the latest commits I feel
      • 2025-06-30 18126, 2025

      • suvid[m]
      • 2025-06-30 18144, 2025

      • kellnerd[m]
        <suvid[m]> "kellnerd..." <- > <@suvid:matrix.org> kellnerd... (full message at <https://matrix.chatbrainz.org/_matrix/media/v3/download/chatbrainz.org/rwBFeYVIphSmEjxVhjonubfE>)
      • 2025-06-30 18100, 2025

      • kellnerd[m]
        If I remember correctly, the LB API documentation is not very specific about whether the duration should be playtime or recording length.
      • 2025-06-30 18112, 2025

      • kellnerd[m]
        The former makes more sense for cumulative listening time statistics, while the latter is arguably a bit more helpful to identify the recording.
      • 2025-06-30 18138, 2025

      • lucifer[m]
        suvid: sure, i'll review tomorrow.
      • 2025-06-30 18159, 2025

      • petitminion joined the channel
      • 2025-06-30 18107, 2025

      • petitminion has quit
      • 2025-06-30 18119, 2025

      • Kladky has quit
      • 2025-06-30 18148, 2025

      • lusciouslover has quit