#metabrainz

Well, if we add autoselect it's either block (with a message that says the links have a short lifespan) or autoselect (with a guideline that says use only if useful) I guess

You think the latter might be better?

Yes

Ok, let's do that

It seems events use IDs such as 00005ED3057A881F so we need to be more specific with them then

Or less, rather :D

reosarevok: About things too obvious to be seen, do you have an idea of why the cover art image is not displayed for “Add cover art” edits with merged release with https://github.com/metabrainz/musicbrainz-serve... ?

Gimme a sec and I'll look

I checked the first example, the cover art image is still accessible from the CAA.

outsidecontext: I use windows to tag files because mac is very very old and the only picard that works on it can't even access th new api so it can't pullreleases from mb, it's 1.4

outsidecontext: right. will make sure never to use "shift" while pulling files to picard (or alt or ctrl either :D)

yvanzo: what do you think about sidebar or not for these btw? Any preferences?

ApeKattQuest: 2.9.1 will be out soon and will fix this

👍

reosarevok: I’m not really interested but I’m pretty sure that it will be soon requested if not added.

I guess

Ok

"Since these pages are often temporary and get removed after the event passes, only link to them if there is useful information provided that might prove useful later (for example when viewing an archived version of the page)."

Something like that for the event-url rel?

👍

Ok, updated that

Thanks!

Will commit the ticketmaster select, then look into your PR

yvanzo: Hmm. you're doing $self->c->model('Release')->get_by_gid($self->data->{entity}{mbid}) and setting the edit data mbid if so

But if that exists, but is a redirect, don't you want to set the destination MBID rather than the one stored in the edit?

(untested)

And the numeric id would be gone in that case?

line 114?

Of AddCoverArt yes

This `field_release` variable is not used to load the image.

Hmm

It is loaded from `artwork_release` instead.

EditArtwork displays the error we're getting if there's no release.gid and that comes from display.release though

Is the gid being passed null?

Wait, no

If there *is* a gid but there's an error

My bad

Thank you

I will check the git history of EditArtwork, that seems to be the cause of it.

*.js

It actually comes from the TT implementation which assumed a GID for removed releases.

yvanzo: the error is shown though AFAICT because it's trying to load https://ia600307.us.archive.org/5/items/mbid-fa... which does not exist

Apparently it's now at https://ia801604.us.archive.org/9/items/mbid-75...

I wonder how the calculation of the prefix works

I see the new MBID should be used instead.

Yeah, at least that

which is weird because this loads just fine: http://coverartarchive.org/release/fabacf46-986...

CAA handles the redirect here ^

Yeah, seems so

But because we're intentionally skipping CAA to save calls...

I expect it's that, anyway?

yes

lucifer: reclaimed 100GB more from Docker on wolf :)

awesome!

zas: I already prepared the blog post for the Picard 2.9.1 release. If you have time, can you look over it (it's in Wordpress' post draft list)

lucifer: has your visa trouble been resolved?

outsidecontext: nope, consulted multiple people. all said nothing that can be done in time to attend this year's summit :/

oh no, that's a bummer :( I had been looking forward to meet in person

such a BS

yup, same here.

indeed :(

outsidecontext, I know you're planning on issuing Picard 2.9.1 tomorrow. Is it worth me trying to add the list of plugins requiring update and a button to open the Options -> Plugins page directly today?

* add to the notification messagebox

rdswift: I think the button in the notification can be definitely still added. not sure about the plugin list (not sure how well this works, especially if there should be a long list)

Okay, I'll take a look this morning. As for the potential long list, I could limit the list to 5 plugins and if there are more then add "plus others" or something like that.

good idea

On it. Thanks.

mayhem: about popularity dataset, i have tried to generate it twice on combined data but it fails. memory issues probably. another run currently underway takes 3hrs to know if it'll work.

ok, if it fails, I can probably do the merge of the two datasets in PG or just straight off the FS.

we will need to examine the j5 / resource question before too long. :(

yeah makes sense. MLHD is a bit for the cluster to handle.

outsidecontext: the blog post looks ok, I did few minor changes in phrasing, and made the part about recommended upgrade for win users more visible.

ugh. now disk issues on kiss.

lucifer: are you current projects possibly causing problems on kiss?

disk wise?

mayhem: no its the full dumps.

ah, mid month.

yup.

I think we really should switch to zstd compression, that will help a little.

should we delete the oldest full dump?

since we're about to get a new one?

sure i'll do that.

ok

listenbrainz-dump-1388-20230715-040003-full

is the one I was eyeing.

zas: thanks

fuck ran out of diskspace.

lucifer: I deleted the files.

oh, site back up.

mayhem: thanks

np