Collaborating to fix messed up recording data

MusicBrainz
, ,
1

Advance apologies, this is very long --if you just want to see the actual question, skip to the last paragraph (but, forewarned, if your answer skips the context, will send ya back to read the whole thing, it’s a simultaneously complicated and niche/lame question but…hopefully someone has a good answer).

Background, feel free to skip: my current work here and for the last few years (though not my only work, current or otherwise) involves a lot of research into much older recordings, releases, and works, poorly represented here despite their importance to (“popular”) music. Details are ambiguous, lots of stuff especially internationally, got very poor rights control, got re-recorded, sometimes on purpose and sometimes just takes, etc. The combination creates real, actual, separate recordings (not in the MB sense, but rather the musicians involved made them multiple times, sometimes not the same musicians). But if it’s NOT diffferent, well, it should be the same MB recording.

TL;DR: Where do you discuss the confusion over some particular work/artist relationship over time vs. what’s in the MB db so far?

More detail of that TL;DR: What would really help is if the people who cared about artist x or recording(s) (titled) y could get together and figure out how to get to the truth together. Instead, what you get is an editing system where you kick off merges, big or small; they get seen and voted, or not; and data debt is created if they’re wrong.

Here’s a concrete example, alas very, very specific, since I’m asking about a general case. This major figure of blues/rock history in the 1940s/50s called T-Bone Walker wrote a song called, with all its various release titles combined, “(They Call It) Stormy Monday (But Tuesday Is Just As Bad)”. He made two serious recordings of it, one at a debated date in the 40s that became THE definitive hit, and a second in 1955-56 for Atlantic, that also did just fine and entered the canon. Plus, many live versions, plus, perhaps some other studio versions. Hard to know what happened, exactly.

So then you turn to the other thing meant to help, the acoustID, and if you’ve spent one minute merging well here you know that’s not reliable. In this particular case though, with many examples here, there are in fact two that seem quite good… but also mixed, since you would expect the Atlantic/Warner family of today (notably Rhino, but there are of course others) to just release their own. But I can find an example of a Rhino release claiming they’re releasing the prior version (for the few relevant nerds, see Recording “Call It Stormy Monday” by T‐Bone Walker - MusicBrainz and https://coverartarchive.org/release/dd191484-e3de-4097-b37c-d25bde1083b4/7152123001.jpg) and NOT their own parent’s one. And, predictably, both IDs in question are on the recording. (Happy to provide, hoping to avoid deep weeds.)

TL;DR #2 and I promise the actual question: The point is general: where can MB deep-diver editors collaborate on ONE artist, ONE work/recording, ONE release/group, in order to make better edits, before making them? And if it doesn’t exist, could it?

5 Likes
2

The following is personal opinion - just say what I see. Don’t mean to annoy people with my comments. :slightly_smiling_face:

The answer seems to be Forum or Discord or IRC - at least three different channels of different communities with a different focus.

My favourite source ends up being the edit notes of an actual artist. You bump into fellow editors who are interested in the same artist. There are a lot of editors out in the edit notes who understandably never visit the “social” channels.

Down side is you start to realise how small a number of editors there really is, so people are spread widely. And each of the separate communities have their own little cliques.

There has been talk of the GUI getting some user tags attached to Artists to say “edited by xyz” so you can see who is most active editing there so you can attempt to PM them. That would be a nice idea. (See relevant ticket)

As to AcoustIDs. Now you talking my area. They can be a mess. So many ways to get bad recordings linked to them, and therefore bad AcoustID data. So one has to tread with care. Sometimes bad submissions from Picard, sometimes bad Recording merges, or just a user adding a new release and linking the wrong Recording.

All we can do is make best of the mess we find. My attitude is when I see a mess I fix it best I can. Try and leave some notes. Some kinda trail to my sources. Attempt to get others interested, but realise mostly this is a lone path. Requesting a pile on of editors is more likely to get lots of people who don’t know the subject.

TL;DR answer - Good Luck.

8 Likes
3

It’s possible to PM/DM other editors? Do you mean the “send email” link on profile pages, or is there some other in-server way to do this that I haven’t come across yet?

https://tickets.metabrainz.org/browse/MBS-9240 requests adding a way to tag/“@” editors within edit notes, which I would personally find useful for contacting past editors when doing cleanup. Right now, I usually just leave notes on ancient edits linking to one of my new edits so discussion can happen there.

4

Yes, I meant the Send Email link. It is an anonymous system and does kinda work. I’ve had a few conversations with people about an artist that way. They can still reply via the same “Send Email” link and keep things private/anon.

Though, to be honest, the mailbox I have linked here I rarely check… I more rely on the message bar as I login so I’d prefer to see a real PM system that would notify me as I login. I’d not see an @tag unless it came up on that bar like Notes left on my Edits. So leaving a note on a five year old edit of mine is more effective than tagging me would be.

What I’d really like to see is some kinda “notes” section too. Was working on some artists just now that are clearly not maintained. Would have been good to leave notes where they can be seen as Edit History is too long and full of other items. A kind of “editors annotation” field for just the editors to see and converse in about that entity.

2 Likes
5

Yes this is more along the lines of what I mean–a kind of mini-wiki/forum on each entity, more formal than an annotation, with no effect like an actual edit, and with rudimentary reply function, only visible to editors, and probably in the vast majority of cases simply empty. No doubt non-trivial to implement but it would allow for simple asynchronous communication and better collaboration avant la lettre.

2 Likes
6

It doesn’t even have to be as posh as a wiki. Just a focus point for comments would be good. Edit histories can soon run to many many pages. A simple edit note specific for the entity would work. Especially as this would then trigger replies to an edit note once you make your first comment to it.

7

Not sure what you mean by this (“edit note” already means something), but I am specifically talking about something visible/accessible from the entity page itself, not any part of edit history or anywhere hard to find.

Not that it matters, this is pipe-dreamin’ with the backlog of features to do…

1 Like
8

I meant it could be kept simple. Something accessible from the entity page itself, but not lost in the edit history. Maybe on an extra tab.

I meant “simple edit note” as in it could just use the same system as edit notes to write the comments. Just somewhere that is always accessible and easy to find and specific to that entity.

Something editors would want to see, but not get in the way of a researcher.