Encouraging Adding Edit Notes

I fear the only way to stop people from adding no or bad edit notes is to block edits from being applied after 7 days if they didn’t get a single yes-vote. On the other hand that would get more people to create fake accounts to vote on their own edits.

I’m going to start voting “no” on all merges with bad edit notes soon.

The code of conduct says:

Never vote against edits where information is optional (not required during the submission process). Instead Abstain and enter an edit note requesting more information.

For merges leaving an edit not IS required and leaving one containing no information is not enough.

This goes for edit notes like “merge”, “same”, “merging into oldest MBID” or “compilations”.

5 Likes

I wrote some code to block edit notes with only spaces, punctuation or one character. That might block some of the bad notes, although I’m sure not all of them.

9 Likes

Isn’t voting no on possible good edits a bad voting practice, though? Perhaps you could try to ask the editor for more information first? I am aware that you are reviewing every merge (and I’m very grateful for that!). Perhaps after https://tickets.metabrainz.org/browse/MBS-10299 is implemented it could be manageable this way?

Additionally, even though the following link is not a guideline, I personally try to base my current voting tendencies on the following article: Editing: Making MusicBrainz better – MetaBrainz Blog , although sometimes instead of politely asking the editor to cancel the edit first I do cast a no vote if their edit makes the data worse, however I always accompany it with an edit note explaining why. Also I can’t approve and fix stuff myself cause I’m not an auto-editor (doesn’t matter for merges in this case anyway).

5 Likes

I disagree. A lack of votes should not be an assumption of wrong data. I am often working in obscure artists doing edits that never get votes. Probably due to people not always knowing much about these artists. Rejecting edits because they are being done in an unpopular area is wrong.

Sock puppet accounts would soon stand out. You would especially notice them as you see so many of the edits yourself. If the same users are always voting for the same users then it soon gets obvious.

I do agree with the frustration of the notes system. I loose count of the number of times I have tried to engage someone in conversation due to bad data or bad edits. Usually starting with Abstain votes to go with a comment. Only kicking that to a No if they clearly break the guidelines or are known bad data.

That looks perfect to me

3 Likes

If “no” votes on bad edit notes instead of bad edits as threatened by @paulakreuzer should become current accepted practice that would be a major change of the guideline quoted and followed by @culinko, which states “give people the benefit of the doubt that they are doing the right thing”.
As much as we wish and ask for sensible edit notes, we also ask for reasonable doubt and comment with no votes: a missing or unspecific edit is never a sufficient reason for voting no
See also Edit #62907424 - Merge recordings

3 Likes

I wasn’t a 100% serious with my last post and also obviously I wouldn’t do this with editors I hadn’t asked for better edit notes many times before. Also obviously my no-vote would still come with a note asking for more info and I’d switch to abstain as soon as they provide some (even if I don’t agree with the info).

Generally I do so too even though I hadn’t read that blog post before.

One thing though. The post says:

Edits fit into four categories:

Edits that makes things better (perfect or not)
Edits makes things different (but neither are better)
Edits that contain some correct things and some incorrect things
Edits that are outright wrong (existing data is better)

That is only true for voters who know everything. For everybody (else) there are also:

Edits that you can’t tell yet whether they improve or worsen the data.

That’s for example true for all those recording merges based only on title, length and artist. Maybe one day the compilation one of the two merge subjects is from will be re-released with more detailed artwork providing evidence that the recordings are or are not the same - but right now you just can’t tell.
In this case you can’t make a judgement on whether the edit improves the data or not, you can only weigh the risks vs the benefits and I would argue there is no situation where you are much better off when you already merged before you knew it was deserved, but you are much much worse off if you merged and then find out it was wrong.

I never said that. But a lack of votes often is evidence for a bad edit note. You’re right of course that with obscure artists you are likely going to get fewer votes, but getting one yes-vote is never really hard with a good edit note.

Today I found such an account. 99% of yes votes cast by editor A were cast on edits by editor B and 99% of yes votes cast on edits of editor A came from editor B, but all I could do was ask them whether they are accounts by the same person and when they said no all I could do was believe them.

I agree with the first part, but I would argue the link doesn’t really fit here.
In that edit the problem is no longer the lack of provided info about what the edit is based on, but the lack of basis for the edit. We asked for info, the info was given, it just wasn’t convincing.

That’s my whole frustration here. There are a lot of editors, even well established ones, who refuse to add good edit notes even if they are asked for them many times and each time they are asked they either get defensive and aggressive and still refuse to give any info or they admit their edit is based on basically nothing and still refuse to cancel if someone voices reasonable doubt.

I believe the following was the only time I actually voted no on such an edit:

https://musicbrainz.org/edit/58867629

The gist of the long conversation there was:

A: “merge”
B: Based on what?
A: Based on nothing.

6 Likes

Totally agree on this one. It freaks me out seeing some of those huge merges where you can see that they have clearly not checked the data, just looking at the names and times. And clear mistakes get made as there are many editions out there of some tracks. I’m working on a cleanup in the coming weeks of exactly this kind of mess where someone has just done huge merges based on name and times. You can tell these merges were not checked as open up the page on the Release Group of that album and there is the Wikidata intro which STARTS with details of how the initial release was out for only a few weeks with different mixes on most of the tracks. :confounded:

(Actually - the times in that example are very different, but it didn’t stop the mergers!)

Some people are more interested in merges than accuracy.

I misunderstood you then. Sorry for that. (Proabaly read the words too literally) Lack of votes is NOT to do with lack of edit notes. I assume that lack of votes in my case is due to lack of knowledge of the voter on my subject.

As you have noticed, my notes can sometimes waffle too much, but there is only so much time available and editors can’t be expected to add every reference to the edit note. For example, I’ll often refer to cover art attached to the Release and I have seen some voters complain that a link wasn’t included! Yet the art is only a couple of clicks away. Those voters need to remember that editing is also very time consuming :slight_smile:

I like when I see your Abstain votes - that tells me I haven’t done something obviously stupid, but you don’t knwo the artist so can’t comment further.

Please tell me you hit the REPORT THIS USER buttons on their account? This kind of thing should go straight up the tree to someone that can suspend the account. It is an area I have often seen strangely lacking here - there seems to be no real interest having an consequences for wrong doing. I love a bit of anarchy, but there needs to be someone with a BanStick™.

Can I vote for this point 100 times over? Some of those people I wonder if they even listen to music, or are just stuck in a database merging streak. They seem just see words and numbers and have no care that some of us are here to listen to the little differences between releases. The example you give there also leaves me confused as to the lack of care.

That example is painful in so many different ways. That editor is making it clear they don’t care about the music, they just want to tidy up some links. They clearly don’t bother even opening the page involved of these artists. You can see that many of them are not even bothering to check the items being merged - too often you see merges going backwards due to the weird assumption that “old” is better than “accurate”. (ARGH!!! You have got me joining your rant with my own rant! We both care too much about the quality of data :wink: )

They will use an excuse of it is “too much effort to check” - well they should not be merging then! It will take me many hours research to untangle the mess that has been caused on a badly merged artist. It will lead to probably a dozen edits, and a whole batch of new recordings. Meanwhile Mr Merger will have run another 500 Mass merges and be all smug they have move themselves up the Number of Edits tables… and someone suggested making this a “game” with “rewards”! If you think it is bad now, wait until people think there are prizes for the numbers of edits…

I have given up trying to deal with those people :frowning: This is seriously one of the Weirdest “communities” I have even been a member of.

4 Likes

Does your code take into account text written by scripts like the Mass Merge script? As noted by @paulakreuzer above it is those merges causing big headaches. Technically they include a comment as the script splatters text… but the comment is meaningless without REAL words.

2 Likes

That is not something that can really be checked on MusicBrainz’s side. It’s something that individual scripts will have to be upfront about themselves. @jesus2099 at least is aware of this:


If there are other scripts that this is lacking from, please report to said userscripts’ authors.

5 Likes

This page actually encourages people to merge based on name, artist and length, but I see it has not been reviewed by the documentation team.
https://musicbrainz.org/doc/How_To_Merge_Recordings

Wrong merges happen easily when recordings of alternative versions/mixes/etc do not have any disambiguation text. For the same reason recordings get assigned to the wrong releases by mistake. To not make a mistake, the only solution would be to add new recordings for every disc - leading to an even greater number of recordings without proper disambiguation, so I can see why users are encouraged to merge recordings based on name, artist and length…

I’m one who will sometimes make large merges, but I’m careful about it. I check every release I’m merging a track from to make sure that there’s nothing about it that would contraindicate merging: live albums, re-recordings, anthologies with multiple versions, basically if it’s anything other than an obvious “best of” kind of compilation I’ll dig into it further, and if I can I add a disambiguation instead. If I can find evidence of multiple recordings of similar length, I also won’t merge. For instance, Patsy Cline overdubs make me very hesitant to merge a lot of her recordings. But if I’m looking at a “hit song” with only one well-known recording, then yeah I’ll merge a bunch of compilations just based on artist, title, and length. The value added by having relationships shared between merged recordings outweighs the risk of merging different recordings in a case like that, for me.

9 Likes

Yes, in the case of one-off “hit” songs your approach allows reasonable confidence.

3 Likes

Is it normal to not being forced to leave edit note on ISRC remove edits, @reosarevok @Freso ?

Well, it was never implemented that way, so it’s not a bug, but we might want that. That said, it’s not really trivial to change as far as I can tell (and it’s the sort of removal edit that is trivial to undo, same as removing an alias, since it’s just a string and all the info is in the edit), so it’s probably not a high priority.

1 Like

The problem I’ve found is the final page is complaining about problems in the note, yet NOTHING is highlighted as a problem. I’m even trying to refer to the Discogs article on the tape. And since it won’t let me complete the page, that means I can’t finish the entry. And since the site doesn’t give me a way of saving a partially-completed entry (or to export it to a file), I have to THROW AWAY all the work I just did. I don’t think I’ll ever bother to enter another item, because it’s quite happy to let you waste all your time and then laugh and say “sorry sucker”.

I see you’ve had a similar problem in September 2018.

Would you be able to make a screenshot of what you see and ideally of the other sections of the release editor as well?
Also helpful could be knowing the web browser you use and if it has been customized, e.g. extensions and settings affecting Javascript or network requests.

It’s possible you are encountering a bug but without more information it’s hard to help you, others had already requested more info back then yet got no reply.

6 Likes

Editting can be very frustrating in my experience.
And yes not being able to save the work done when it has taken a long time adds to the negative feelings.

I finally realise that I can cut and paste the Track info into a text document.
And take a screen shot of the first Release info page.
Which will save most of my work but not refund the time.

Editing does get easier with practice. Though a few months off and it gets more difficult.

If you can provide screenshots then us volunteers here will almost certainly get things solved.

3 Likes

Note you can programatically seed the release editor - Development/Seeding/Release Editor - MusicBrainz Wiki - if you have a structured data source that you’d want to fill in the release editor fields with.

I wish the same could be said for the release relationship editor…

2 Likes

“This page documents a list of parameters you can POST to the release editor to “seed” it.”

What does POST mean / entail ?

POST is a HTTP request method which should be used to submit the web form with the data you want to seed the release editor with. Have a look at the source code of the seeder example and you’ll spot it:

<form name="form" id="form" action="/release/add" method="post">

So it’s just a small technical detail :wink:

4 Likes