I'd try asking again tomorrow. Or asking on the forum.
2019-03-10 06913, 2019
pac23 joined the channel
2019-03-10 06921, 2019
Freso
But only getting MSIDs and no metadata does sound like a bit of a bug, but without more information on what you're doing and experiencing, it's hard to say for sure.
2019-03-10 06910, 2019
Zastai
I'm just working on my C# lib for accessing the LB API. Using my own imported scrobbles for now (I don't have any players that submit to LB directly I think).
2019-03-10 06929, 2019
Zastai
If you know a user who does, I can try testing using their listens.
2019-03-10 06918, 2019
Zastai
(Just tried yours, and that's MSIDs only being returned as well)
one interesting thing about last.fm, at least last I checked, is that if you send an mbid in the scrobble, it disables a bunch of the track name/artist name filtering & cleanup
2019-03-10 06938, 2019
kepstin
e.g. you couldn't scrobble tracks by the artist MOSAIC.WAV unless you included an mbid :)
2019-03-10 06930, 2019
Freso
MusicBrainz is truly magic.
2019-03-10 06937, 2019
kepstin
but yeah, I don't send listens directly to listenbrainz yet, i still use last.fm imports.
2019-03-10 06930, 2019
kepstin
i got into the whole musicbrainz thing because I wanted better data in my last.fm history :)
(AFAIK, the listens are only linked to recordings if they contain MBIDs.)
2019-03-10 06930, 2019
kepstin
that's weird, because most of those were submitted via simple last.fm scrobbler on android, which you just said doesn't send mbids
2019-03-10 06950, 2019
Zastai
ah yes, recording mbids are set in kepstin's recent listens
2019-03-10 06903, 2019
Freso
Well, it doesn't send MBID for *listens* directly to ListenBrainz.
2019-03-10 06919, 2019
kepstin
ah
2019-03-10 06930, 2019
Zastai
none of the other mbids are set
2019-03-10 06944, 2019
kepstin
yes, last.fm only has recording mbids with any reliability
2019-03-10 06904, 2019
Zastai
ah, and I see that at least some of my recent ones do to; i had been working against a set of last October
2019-03-10 06930, 2019
kepstin
(they return artist mbids in the api, but those aren't from the scrobble - they're determined by last.fm based on an old musicbrainz dump and they incorrectly merged artists in the same name)
2019-03-10 06931, 2019
Zastai
shame that spotify's scrobbling doesn't pass the spotify id along either it seems
2019-03-10 06909, 2019
kepstin
recording mbids are only returned in the last.fm api if they were sent in the scrobble, afaict.
apparently my config is actually sending *track* ids, not recording ids
2019-03-10 06945, 2019
Freso
Zastai: Listens directly from Spotify (using the Spotify integration in LB) should contain Spotify IDs.
2019-03-10 06951, 2019
Freso
Zastai: Check "rob"'s listens.
2019-03-10 06934, 2019
kepstin
(that's probably a player bug on my android player, it should be sending recording ids to last.fm)
2019-03-10 06950, 2019
Zastai
ah yes, rob's listens have more info
2019-03-10 06900, 2019
Zastai
including yet more undocumented fields in additional_info
2019-03-10 06907, 2019
Freso
kepstin: Yeah, lots of people didn't get the memo that MUSICBRAINZ_TRACKID is Recording MBIDs… and some people want to make life harder for themselves/contributors too…
2019-03-10 06929, 2019
Zastai
APIs should be document-first, not code-first, dammit :/
2019-03-10 06941, 2019
Freso
Zastai: You can submit pretty much anything you want in `additional_info`.
2019-03-10 06947, 2019
Freso
Which is also documented.
2019-03-10 06901, 2019
Zastai
joy
2019-03-10 06912, 2019
Freso
"Add additional metadata you may have for a track to the additional_info element. Any additional information allows us to better correlate your listen data to existing MusicBrainz-based data. If you have MusicBrainz IDs availble, submit them!"
2019-03-10 06935, 2019
Zastai
additional-info.line-1: "It was a dark and stormy night." :D
2019-03-10 06920, 2019
Freso
And yes, it is a joy. It allows users to send any kind of identifiers etc., without requiring explicity support in ListenBrainz.
2019-03-10 06904, 2019
kepstin
huh, "At this point, we are not scrubbing any superflous elements that may be submitted via the additional_info element. We’re open to see how people will make use of these unspecified fields and may decide to formally specify or scrub elements in the future."
2019-03-10 06955, 2019
Zastai
sucks for people wanting to make nice and clean APIs in code. or rely on any information being there under an expected name in any expected format. allowing freeform is for messybrainz, not for LB, as far as I'm concerned
ok, with some help from the gmmp dev to confirm some stuff, i'l like to let you know that last.fm themselves are putting track ids into the mbid field in their api response!?!
by the way, it it a bug or intentional that recording_msid is returned at the top level of a listen, and not alongside all the other _msid and _mbid fields in additional_info? or inside track_metadata at least?
2019-03-10 06908, 2019
kepstin
I guess I should file an issue for this, but I'm not sure what to say. "MBIDs returned from last.fm api may be either track mbids or recording ids with no way to distinguish them, depending on whether they were submitted or internally determined" is about it
MBS-10069: "has remixes:" shouldn't appear on a release page
2019-03-10 06901, 2019
CatQuest is ammused becasue he's been saying this for years. but now it's gettri traction and suddenly *everyone* is saying it and't awesome !
2019-03-10 06904, 2019
ahmedkrmn_ has quit
2019-03-10 06952, 2019
kepstin
CatQuest: this is actually a slightly different issue, this is a problem with work level rels being loaded. https://tickets.metabrainz.org/browse/MBS-9419 is more specifically for this issue
2019-03-10 06952, 2019
BrainzBot
MBS-9419: Don't show versions, medleys, etc for a work on the release page
2019-03-10 06911, 2019
kepstin
and I filed MBS-10038 with some ideas how to improve it, but :/
