#metabrainz

More entity naming?

reosarevok: I thought it was yesterday?

Oh. Was it? :D

Dunno, didn’t have any answer about Wed/Thu... And it seems that you already made the schema change discussion on Tuesday...

Yeah, because we had to put out the blog post last night at the latest

But I thought we were going to have the meeting about further naming issues today :D

Maybe not

I mean, I'm up for it if you and/or bitmap are too, anyway, but if not, no worries

Let’s plan next meeting at least.

Ok :) Any preferences?

Tue/Wed work for me.

akshaaatt: uh they dont?

I just found bitmap's answer about today's meeting in the backlog: https://chatlogs.metabrainz.org/libera/metabrai...

all open PRs have failing tests

Yeah, that's why I thought we'd meet today, heh :D

Tue works for me too, I think :)

Let's see what bitmap thinks

Is there anything you want to talk before that? Or that I can help with short-term? :)

Yes yvanzo, you're right, just clicked in progress. As always, thanks for the help with SEC tickets :)

!m monkey

You're doing good work, monkey!

sorry reosarevok yvanzo my internal clock is still messed up from the DST switch, I can meet in 30 mins if you'd like

bitmap: no problem, I was not even expecting a meeting :P

But ok with meeting at 16:00 UTC :)

[listenbrainz-server] 14MonkeyDo reopened pull request #2173 (03master…dependabot/npm_and_yarn/jsdom-and-jest--removed): Bump jsdom and jest https://github.com/metabrainz/listenbrainz-serv...

Gah !

Yet another open source project were the lead maintainer is no longer able to continue the development in their free time...

It looks as if the Knex.js maintainer was willing to continue doing maintenance releases but there were no status updates for a few weeks.

Better to find that out now though.

So what are our alternatives? Wait to see how objection fares? TypeORM? Prisma? Any other TS-oriented ORMs you know?

hey, so i'm looking for this: https://musicbrainz.org/release/4fed047c-cfe2-4...

Hmm, I already had a look at those (and others), and they're much heavier ORM frameworks whereas Objection is more like an enhanced query builder with support for lazy loading.

and since the boot i downloaded said it was an EC boot, i started there. then i looked at the cover and it's actually a BB boot.

so I went to BB and switched the view to Alll release groups

then filtered

and it the excluded unofficial release groups so https://musicbrainz.org/release/4fed047c-cfe2-4... is not displayed

why?

is that by design or an oversight?

kellnerd: I see. Yes, that would be ideal, especially in view of our ORM issues

Yeah, personally I'm not really a fan of these full blown ORMs and originally intended to go just with Knex (for my personal project) before I discovered Objection back then.

[listenbrainz-server] 14MonkeyDo merged pull request #2396 (03master…dependabot/npm_and_yarn/braintree/sanitize-url-6.0.1): Bump @braintree/sanitize-url from 6.0.0 to 6.0.1 https://github.com/metabrainz/listenbrainz-serv...

bitmap, yvanzo: so, meeting or no meeting? :D

what's on the agenda, entity naming?

yep

I suppose I should address the comments in the Flow PR and we can merge that

then continue looking at other CoreEntityT uses?

monkey: But I don't know whether that makes sense for BB. The ideal outcome would be that Objection continues to be maintained, IMO. Maybe there's still hope, according to https://www.npmjs.com/package/objection kibertoad has accepted the NPM  invitation at least.

There is hope

(Maybe we should move to #bookbrainz now)

Sure :)

iconoclasthero: it shows up in the latest page: https://musicbrainz.org/artist/dcb03ce3-67a5-4e...

Please go back and reread what i wrote.

That is not correct

it doesn't show up.

you need to filter after selecting all release groups

and then it doesn't

But the filters don’t mix with showing all RGs.

yes, that was my question.

why, and is that by design? if so, again, why?

It’s not deliberate, it’s just a rough implementation.

because if it isn't by design, it's implemented wrong

where do i put in tickets?

MBS, improvement suggestion, thanks!

IIRC I have a PR for that?

iconoclasthero, yvanzo: is that what the discussion was about? :)

yeah, iconoclasthero, it's https://tickets.metabrainz.org/browse/MBS-8186

MBS-8186: Extend release group filtering to all sets of RGs, not just official by this artist

okie dokie

thanks

added you to followers

thanks

reosarevok: yvanzo: should we try to resolve the discussion in that PR today? :)

yes

added a list of new types to the PR's description

bitmap: I thought that we would refine types depending on components’ needs (that is specifying need properties, what I tried to do with MergeLink) rather than hard-coding a list of entity types (as in MergeableEntityT/MergeableEntityTypeT).

in that case I think it made the types a lot more complicated, because we needed the component to accept a type parameter IIRC

I agree we don't need a list-of-types type in all cases

the list-of-types is mainly useful if you need to refine the type to a more specific one using the entityType property, so that you can access other entity-specific properties.

TBH, if there was an option to filter by unofficial release groups, that would make my entries of bootlegs easier

@yvanzo

iconoclasthero: See MBS-12460 and MBS-12047

MBS-12460: Allow filtering artist releases by status https://tickets.metabrainz.org/browse/MBS-12460

MBS-12047: Support browsing release groups in /ws/2 by official status https://tickets.metabrainz.org/browse/MBS-12047

12460 seems to be aiming to do the opposite of what i want?

unless you can filter the other way too

it's aiming to allow both, excluding bootlegs is just an example

k

I don't understand what 12047 is ssaying. what is /ws/2? what does it mean by browse request via website

?

It's to allow to do the same through the API, not the website.

regardless, the functionality for entering bootlegs isn't considered in the OP

I mean, just pick any random dead bootleg from archive.org and see if you can figure out if it's in there or not.

(OP)

?

isn’t it already possible to enter bootlegs?

lol...

you've obviously never tried to do what i just said.

no

is this in MBDB: https://archive.org/details/gd1987-12-30.schoep...

?

that should be since it's an oakland show around NYE

though prehaps not the etree 84986 version

Since until recently etree wasn't an option for label number

there's also no good place for torrent hashes or names

i usually put them in the disambiguation

reosarevok: yvanzo: not sure what to do about the CoreEntityT usage in e.g. EntityHeader. since that uses EntityLink, it will have to be a union type. so, we could use RelatableEntityT, but it's not clear that EntityHeader is only for relatable entities.

Well, we should probably use it for whatever we expect to use it

So if it's only Relatable right now, use that, if not, make a union type?

sure. RelatableEntityT works fine, so I may be overthinking it

Here's a 24-bit version of that same show: https://shnflac.net/index.php?page=torrent-deta...

etree id 120547

iconoclasthero: it isn't in MBDB

if you found it easy to figure that out, please tell me how you searched for it

because it is a PITA for me

no it isn't

there's a 12-30 from 1979

and the 87-3-3 show

no 1987-12-30

we agree :)

BTW, look who entered the last entry

that has an etree id and the name of the boot

1990-12-31

oh it ISNT

sorry, misread the highlighted text

i guess the filter function is way easier than going to each page and hitting ^F and searching!

an advanced search is another option at the moment: https://musicbrainz.org/search?query=artist%3A%...

you will still have to ctrl+f, but if it exists it should be on the first page

bitmap: Would it be possible to generate TaggableEntityT type declaration from entities.mjs?

is there any way that the entry box for annotations could be expanded rightward?

We have had issues with mismatch between entities.json and Perl code, we don’t want to have it again in JS code.

it's technically possible, but no code exists to generate types like that yet

iconoclasthero: It is possible in the annotation editing form that can be accessed from “Edit annotation” link below the annotation display.

but, if you tried to use SidebarTags or TagEditor etc. from an entity page that isn't in TaggableEntityT, it will error. so they shouldn't really silently become out of sync

yvanzo is that only after the entry's been submitted the first time?

iconoclasthero: yes

I don't know what those Tag things are so moot.

It's a different topic as we are having our weekly MB dev meeting atm.

ok, later

It still is code duplication.

It might be possible to add a type to Taggable*T that is not in entities.json.