#metabrainz

Even resetting the timestamp didnt create dupes.

Did you import the listens after the artist mbid pr that I made

I imported just now

For the first time?

And my lb profile had listens before that PR too

ishaanshah[m]: Naah. I import daily

If the issue reported by Zastai is true. My profile should have at least 800 to 1000 dupes

ishaanshah[m]: Did you give it a try?

Hmm, I am not sure then, I was getting duplicated listens for listens which were imported pre artist_mbid pr

So I deleted all the listens, imported again

Then even after reseting and importing again I was not getting duplicated listens

Seems like we need something like "Delete my Last.fm listens"

Yep

That would be useful

might be tricky since older imports don't have listening_from to help identify them

I'd settle for "delete any listen not from spotify" tbh

because those aren't recoverable AFAIK

just for reference, this query returns two pairs of my dupes. wonder why the messy id would change if the main identifiers are the same (unless the artist mbid is now factored into the msid)

https://beta.bookbrainz.org/revision/13113 This doesn't look good to me. Almost impossible to figure out relationships

Mr_Monkey: I am planning to do it like this

What do you think?

CatQuest: Any suggestions?

Zastai: messybrainz takes the entire listen json into consideration when generating messybrainz IDs

:/

prabal: my first advice is to look at how musicbrainz edit histories look. take ques fro mthere

but

!m prabal for going to make revision history better in the first place

You're doing good work, prabal for going to make revision history better in the first place!

MusicBrainz for Android has been merged into Fdroid Repository. It should be visible from Monday.

Thankss

_lucifer: 👍

[listenbrainz-server] paramsingh opened pull request #789 (master…param/spotify-player-typescript): Convert playback-controls to typescript. https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] paramsingh opened pull request #790 (master…param/frontend-test-sh-improvements-1): Add ability to update snapshots to frontend-test.sh https://github.com/metabrainz/listenbrainz-serv...

!m _lucifer

You're doing good work, _lucifer!

I stopped scrobbling to Last.FM a while ago, so I’d be losing a lot of unique listens if I dropped my Last.FM history on LB, but I am also interested in re-importing to get the changed/improved tags. It would be nice if there could be a tool to compare potential duplicate listens and you could decide to merge or delete them. (For people with a long history, it might take a while to get through them all manually, but I would

personally prefer being able to go through them that way.)

_lucifer: Seems like I’ll finally get to install it on my phone then. :)

Thank you!

*if I dropped my LB history

[listenbrainz-server] ishaanshah opened pull request #791 (master…recent-listen-port): [WIP] Port RecentListens to Typescript https://github.com/metabrainz/listenbrainz-serv...

iliekcomputers: I left a review on your spotify PR

same as freso. I would definitely want to go through my history and "fix" or "check" things

iliekcomputers: ping

Pong

I just read your comment on my PR, I agree with keeping the props and states in the same file

Great!

Can you just move the spotfiy direction type to types.d.ts

Because that is needed in RecentListens too

Yeah, I was gonna do that, sounds good.

I'm probably not gonna work on it today, however

Hi all. I'm encountering some issues with MBP 2.3.1

Yep, no problem, I am nearly done with porting RecentListens, I will complete it by tomorrow

Great, sounds good. I also am having some problems with getting Spotify permissions to work in my local environment

Adding music that's already been tagged by MBP seems VERY slow and more than that, music that was tagged correctly by MBP now show up as [non-album tracks] with certain tag metadata being really weird.

Looks like we'll have to test it on beta

The migration is not very risky because we're just adding types most of the time.

Sound good

A snaphshot test for RecentListens should be fine for now right

Yeah, just make sure to put in a list of listens

:D

(ideally with mbids and Spotify ids)

Just try to get as many cases as possible into the the snapshot

With mbid, without mbid, with Spotify id, without etc etc

Ok, will do

SkweezyJibbz: You might want to search the forums for related topics. Picard is known to be slow in certain conditions, which have been iterated multiple times there.

Freso: i can do that for sure. but beyond the slowness, MBP is not recognizing songs that it's previously already tagged

it's saying they belong to [non-album tracks] as opposed to the artist/album it already tagged it with

What have you set it to ignore MBID when reading new files?

*Have you set …

No

That option is left unchecked

And the files do have MBIDs?

A lot of them show up as this: https://imgur.com/a/5usZ8ML

Yeah, so that’s a no. Not sure how the MBID tags got mangled up.

That seems to be the case for a LOT of the files

Totally mangled tags

Pretty sure Picard wouldn’t have done this, unless you have some rogue script/plugin going on messing with this.

Nope. I have never downloaded any extra plugins or scripts.

It's stock Picard.

Then it’s likely something outside of Picard.

(Filesystem corruption?)

Looks like some MBIDs are preserved anyway, so it might be possible to recover via a script.

I’d suggest that you post about it on the forum and maybe someone there can help you more.

I can't see how. I'm running a practically brand new version of Windows 10 fresh install.

Yea sure, I can do that

Harddisks (and other hardware) can still be faulty on a fresh software install.

[musicbrainz-server] reosarevok opened pull request #1458 (master…MBS-10756): MBS-10756: Remove link_order from rels sort https://github.com/metabrainz/musicbrainz-serve...

MBS-10756: Inconsistent default sort order for recordings on Works page https://tickets.metabrainz.org/browse/MBS-10756

I'm guessing I should post in the MBP sub-formu?

sub-forum*

Yes, please. :)

[listenbrainz-server] paramsingh merged pull request #790 (master…param/frontend-test-sh-improvements-1): Add ability to update snapshots to frontend-test.sh https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] paramsingh merged pull request #789 (master…param/spotify-player-typescript): Convert playback-controls to typescript. https://github.com/metabrainz/listenbrainz-serv...

SkweezyJibbz, is it possible that something in a tagging script is mangling the MBID tags? That would only be the case if you have a tagging script enabled. Note that the renaming script won't affect the tags written to the files.

rdswift: I only have whatever the stock setup is

Finally, what type of files? (e.g. MP3, FLAC, etc.)?

(These are the questions the devs are going to ask when trying to figure out what might be causing the problem.)

deployed more typescript to beta, the spotify player is now partially typescript

the linter also pointed out some accessibility bugs and now the player has keyboard support as well

rdswift: only mp3 and m4a

That shouldn't be the problem then. As far as I can tell, both are fully supported.

Well, created a post: https://community.metabrainz.org/t/corrupted-mb... so hopefully we get somewhere with it :)

how many files are you doing at a time?

extreme slowness could be a case of lots of files selected, plus it might be rewriting tags (say v2.3 as v2.4); that might also yield "corrupt tags

Zastai: i'm doing about 1,800. have always used v2.3

"corrupt tags" in other software because it expects different tags for a particular property

yeah then the "extreme slowness" is possibly to be expected. picard is not intended for mass tagging like that, more for one release at a time. a huge batch will make it seem unresponsive. it should still be doing its thing though

as for the corrupted tags, I have no idea

I only rip to flac

Zastai: i thought it was doing it's thing as well, but I left it running for like 10 hours and I think the program just froze up completely.

yeah ten hours should have been enough for 1800 files

So, I can circumvent that issue relatively easily by just selecting only 50 files at a time I guess. But the corrupted tag issues is a whole 'nother thing

yeah, batching in groups of 50/100 should work around that part. but not the tags

what corrpted tags exactly?

see the discourse post. values being substrings of tag names

[listenbrainz-server] paramsingh opened pull request #792 (master…param/spotify-player-tsx): Convert SpotifyPlayer to typescript https://github.com/metabrainz/listenbrainz-serv...

Mr_Monkey: if you get the chance, can you look into this: https://tickets.metabrainz.org/browse/LB-521 ?

LB-521: Spotify player does not work on Chrome