#metabrainz

we should in the future have an API endpoint where we provide a list of recording_mbids and get back that list of recording_mbids with last (or never) listened at.

the thing is, all other patches will need this too.

well, not so much new jams, but artists radio does.

i think TS can support that fast once we resize the chunks.

scan the 8-10 latest chunks only.

it would be nice to not have to make yet another dataset.

that gives us whether the track was listened to in the last 3 months

that should be sufficient.

I forget, we still had a rewrite of the listens table planned, but that hasn't happened yet, right?

i'll add that endpoint and once we resize the chunks i hope it to become faster.

resizing chunks should be done at the same time?

not necessarily,

but we planned to do it at the same time.

the rewrite is pending on figuring out how to display the timezone data to users.

ok, sounds good. that will improve the run time of the upcoming patches. they keep getting more and more intensive. :(

how to display it or how to store it?

figure out how to display so that we can work out what exactly to store.

what are the sticking points on this topic?

LB-1213 is the ticket.

LB-1213: How to display timezone aware listens to users? https://tickets.metabrainz.org/browse/LB-1213

but i see it doesn't have the updated info from the last discussion with aerozol and alastairp. i'll re read the chatlogs and update it.

and we can discuss it again sometime later this week i guess.

ok, let make a decision on how to proceed this week.

[listenbrainz-server] 14amCap1712 merged pull request #2381 (03master…dep-updates): Update numpy and decode-uri-component https://github.com/metabrainz/listenbrainz-serv...

cool sounds good

[listenbrainz-server] 14amCap1712 merged pull request #2383 (03master…ac-mbid-fmt-fix): LB-1223: /1/metadata/lookup should return array for "artist_mbids" https://github.com/metabrainz/listenbrainz-serv...

[listenbrainz-server] 14amCap1712 merged pull request #2382 (03master…current-pin): LB-1232: Add an API endpoint to get a user's current pin https://github.com/metabrainz/listenbrainz-serv...

monkey: how do you center an image in a div? lolololol. rookie question. :)

can I put the explore page back on test for you to look to give me some tips?

Ahaaa, it really depends.

Sure thing, you can put it on test

I've spent half an hour on it and I am not making any headway.

ok

Is that the background image that is clipped?

yes.

it should be vertically centered.

Then you're looking for this: https://www.w3schools.com/cssref/pr_background-...

thanks

Most likely `background-position: center center`

does background-position require the image to be CSS backgroun image? because right now its an img tag.

monkey: https://test.listenbrainz.org/explore/

Ah, I see, that's why I was asking if it was a background :) It won't work that way then. Let me take a look

oh, sorry, I didn't quite understand that question correctly.

No worries.

CSS, amirite?

yerp

OK, maybe not the absolute most bare-bones approach, but definitely the easiest way to do this is to add the `flex-center` class to the parent element.

This is one of our custom classes that makes the parent a flex container and centers the children.

I reckon that does what you had in mind

parent being explore-card-img-clip?

yes!

yes

what am I doing 'wrong?

I have no idea. Works in my browser… What browser are you using?

Nope, works in FF https://usercontent.irccloud-cdn.com/file/XX96y...

FF. odd.

🤦

looking at the wrong tab. FFS, Mr. Kaye.

sorry.

Hehe

monkey: ^^ I think changing the breaks is the only thing that is left. any tips on how to do that?

this is bootstrap right?

Yes, media queries. But this means departing from bootstrap scaffolding

I'm trying out some of the existing BS breakpoints to see if they could match those dimensions close enough

ok

can you not use bootstrap breakpoints?

ah

Other wise it will mean removing the BS scaffolding classes, and writing custom CSS to manage the element width as percentages depending on screen size. Not toooo ticky.

is that worth it? for this page?

It's only a few lines of CSS in the end

ok

zas: oops, I just noticed you replied only to the account admins account with that French email

Because for some reason reply-to is kept when they answer me?

I'll forward it to them

We're too limited by BS breakpoints I think mayhem.

Instead of using BS scaffolding classes (col-xx-xx), you'll need to give the card container element a class name to target.

Then here's what the css would look like for those elements:

Oh, and you will also need `float: left;` to replicate the way bootstrap grid elements stack neatly

And apply the padding again as well :)

`padding-left: 15px; padding-right: 15px;`

sorry, I can't say I am following all of that.

Yep. One sec plz :)

So: the top-most divs for each card get their bootstrap classes removed, and replaced by a custom class name. I went for `explore-card-container`: https://usercontent.irccloud-cdn.com/file/F7Fzm...

And then here's the css code for said class:

You'll see we defined a width of 100%, then in the media queries below we overwrite it based on the screen size

In LESS we can shorten that a tiny bit because we can do nesting:

what is with the `::after` in https://usercontent.irccloud-cdn.com/file/F7Fzm... ?

not working flawlessly here:

let me push to test

I must've gotten something wrong: https://test.listenbrainz.org/explore/

mmm, something not quite right with one of the card images, let me dig

Right, the cover art collage image when resized is smaller than the 171px we need to fill the card image container, which results in this weird layout. one fix is to change the .explore-card-img-clip `max-height` css property to `height` instead, which will fix the layout (but not the image ration as you can see in this screenshot:)

confirmed.

That was all probably introduced by the flex centering of the images

If I'm honest at this point the easiest fix is probably to resize that one image…

the collage one?

Yes

k, let me try

[musicbrainz-server] 14reosarevok opened pull request #2882 (03master…MBS-12962): MBS-12962: Don't ISE on undef track_offset https://github.com/metabrainz/musicbrainz-serve...

monkey: thanks for all the help, let's hope that was the last of it: https://test.listenbrainz.org/explore/

Oh, I'm sure i'll come up with some pesky details in review ;p

That being said, it looks great at the moment

!m

!m mayhem

You're doing good work, mayhem!

Stamp of approval as far as I'm concerned

let me get the PR ready for review then

monkey: I merged in the latest master but frontend tests are failing from... unrelated things?

It looks like the linting of your less files is what's failing, unless I'm missing something. Try running `npm run lint:less` if you have node installed locally, or `./test.sh fe -f` for docker version