I think that most of my points have already been voiced here. I’ve put some general headings to break this out as this will be long.
Yes I agree playlists made by your mate Dave on Spotify and made public shouldn’t be in the database.
Yes I currently agree playlists made by Spotify, artists or similar large entities like record labels shouldn’t be in the database as a release [examples at bottom of post].
I personally think that a digital bootleg is a release, regardless of its name (many of them will have semi-official sounding titles to reel people in). I don’t think these bootlegs are as popular as they were back in the heady days of blog and forum popularity, and for those I think there is some digital archealogoy interest present. I’ve (surprisingly considering my tastes) never really bothered with indie/rock compilation bootlegs but have downloaded plenty of “EDM” variants which can include ~100 mp3 tracks of various states of legitimacy - they may not explicitly have the word playlist in them, but even if they do I’d still consider adding them as a release and setting the status to bootleg.
Muting notifications is an avenue to explore
I think the ability to mute content notifications from particular “avenues” might benefit some here, as you could state that anything from a particular label. Although I’ve never seen such a notification email, I would hazard a guess a stop-gap would be to use the automatic rule filters available in most popular email services to just read and delete when there’s a match against a phrase.
Approaching what should and shouldn’t be recorded in a level-headed manner
I don’t want this to get into the sticky, and rather harmful, opinions of “well this has no historical relevancy to ME thus it has no historical relevancy to ANYONE” - that’s not very fair, I personally might not give two hoots about a slightly different recording of a Pink Floyd, Bob Dylan or Grateful Dead concert but I know that the big fanbases these artists attract DO care. I know everyone approaches MBz differently and has different motivations and objectives of what they want to get out of the project, but if you’re really dedicated to “the cause” then a level of ambiguity towards what data is submitted has to come into play. That means not only big mainstream releases are covered, but so are DIY CD-R’s from a one-man-band who plays in your local town hall, so are teenagers making experimental music in their bedrooms developing new genres and styles, so are physical and digital shovelware releases for musical detritus like panpipes, meditation music, sound effects, library music and royalty free music. I know there are a few people who get mad and pee their pants when their favourite musicians release isn’t covered on MBz but to them I say “go forth and add it”. Everyone here is spinning in their own circles, the platform is a way of bringing those circles together to produce a coherent product.
Using the actual playlists as evidence for merging…
These releases are problematic due to their name and the fact they have been released with a playlist link; I can tell you (as someone who participated in the art himself) that there are thousands of playlists on these platforms that mirror compilations that aren’t on the platform in their own. These playlists of course are useful to the listener but should not be used as a method of working out “matches” as they will simply go for the most popular version (i.e. the download might include a particular mix or edit that Spotify doesn’t have, so the uploader will compromise with the album version, which they do thave).
Trying to keep the peace and working out solutions to this
The MB-sphere already has ways to record playlists, ListenBrainz has the functionality at the moment and I’m sure some aspiring nerd among us could probably script something to import playlists from Spotify. Now LB’s playlists are considered personal to the user, but I’m sure there would be ways of making them “public”.
Of course that could end up with duplicate playlists, and doesn’t really serve as a historical document - which is sort of one of the base goals of MBz - to be a record of what was released, when, what was on it etc.
So then the question comes is there value in creating a new entity type called Playlist in MBz with the idea that it captures these problem entities without trying to force them into the established release entity type. Maybe, but then where does that end - do we say “only playlists curated by XYZ are allowed” or is it a “anything goes” approach (again see previous points about the difference between “of importance to me” vs “importance to others”).
You could argue we do have the ability to create lists within MBz already; either as collections (which I get the general feeling casual users don’t know exists, or find lacking in functionality) or as series (which can end up looking rather messy as these relationships will appear on the track entities themselves).
I don’t think there is a be-all, end-all solution to this issue as it currently stands. And each general entity type should be assessed on a case-by-case basis, although one could argue that if a download for it is available (be it legitimate or not [its usually not]) then it should be a release.
Examples of “Playlists”: