#metabrainz

Thanks for all the editing too! :)

Thanks ListMyCDs. Take care of yourself though and don't let your MB project stand in the way of your health. :)

+1

Neither Leo_Verto nor alastairp or anyone else has said anything, so let's carry on.

Freso: Google Code-in

Not much to say. People are slowly starting to formular actual tasks now, though they haven't been moved over to the GCI interface yet. I will hopefully have some good, long days in the MeB office this week, looking over, adding, and editing tasks and getting caught up on our GCI stuff.

fin.

Freso/chhavi_: OTHER-310 and Design Kanban Board

OTHER-310: Add Jira workflow for design/UI/UX/etc. tickets https://tickets.metabrainz.org/browse/OTHER-310

I'll get to adding some tasks today / tomorrow ideally too

I'm all tasket out

I can speak for chhavi_ 's part

So I did https://tickets.metabrainz.org/browse/OTHER-310 which means almost all projects in Jira now have a "Design" issue type which has its own workflow, slightly different from the regular ticket workflow.

and I ended up making a Kanban board for chhavi_ to allow her to manage her design tickets and also provide a pin board of sorts for people to look at active design tickets and provide feedback

I'd like people to keep an eye on this and let me know if anything else has gotten disrupted because of it, or if you think of some ways it can be improved. I've left OTHER-310 open for now, so comment there.

I'll close OTHER-310 in a week or two once I'm sure no one is complaining a lot. :)

OTHER-310: Add Jira workflow for design/UI/UX/etc. tickets https://tickets.metabrainz.org/browse/OTHER-310

Kanban is supposed to be used in an agile env. with rapid prototyping/feedback

what *is* kanban exactly?

chhavi_ said «I just wanted to ask, how should I ask for feedback on specific tickets. Should I ping people personally, or have a bi-weekly design slot/meeting where I run the team through tickets which need feedback?» and I'll let samj1912 take it from here.

basically that

CatQuest: see this https://tickets.metabrainz.org/secure/RapidBoar...

yes i saw

I still don't *get* it

so its like a board where you can drag and drop stuff to change status, prioritize tickets, and have a look at which ones are being worked on, which need feedback etc

oh

like you can see there are 4 tickets in the needs review section

that means chhavi has made mockups for those 4 tickets and is awaiting reviews from ruaok Freso reosarevok zas you and other editors involved in the UI/UX changes

right

once you leave feedback you can drag it back to needs mockup and she will start working on it again

you can also order the tickets to give a priority list of sorts for her to work on

and she can order the needs review part to prioritize tickets which are blocking her on reviews

uff da

The to - do section is the initial part where the community can decide whether to work on a specific ticket or not

basically the point of all of this is chhavi_ wanted to know if this was a better / comfortable model to work with than her pinging everyone asking for reviews

thoughts?

We talked a bit of a while ago about having bi-weekly or such UX meetings.

i like the meeting idea

it's easier to give feedback on something concreete

I'd prefer to not do that until after GCI, but I'll be here. OTOH, feedback should ideally go on the tickets. It easily gets too unwieldy to have it on IRC.

yes

after gci deff

also lots of re-explaining the same things

So maybe just a weekly or bi-weekly regular-meeting agenda item where she lists the design tickets she would like to get reviews on?

+1

Ie., not a discussion agenda item, just an information one.

I like it

and to leave comments on the ticket?

yea

+1

good idea

Freso: basically, all the to-be reviewed tickets will be on the kanban board in the priority order

ruaok, reosarevok: ^ ?

samj1912: Sure, but I will not remember to check the board on any regular basis. :/

what Freso said

(Unless I get e-mail notifications or some such.)

and then it can just be like "bi-weekly "these are the changes i want feedback on" anywa

y

I think you can turn on notifications for the board

board still requires going ot the individual tickets anyway

anyways

I think this agenda is getting strched

we can do it after the meeting?

I don't mind whether we do it every week or two or if chhavi_ shouts at me about checking the stuff

+1

:D

Or if I get reminders from the board

But I'll need to get reminders one way or another

Fine with going on later :)

i forget email

Alright. So no resolution to this question. :)

CatQuest: I think it's your turn

(For now.)

yes

it's much easier to get people to "dive in" and learn what does what, when they don't have to worry about breaking anything.

CatQuest: Get test.mb up before GCI.

this is true regardless, but especially for gciers

:D

bitmap, yvanzo, zas: Do you think this is doable? I have no idea what current status is.

samj1912: :)

yes, it should be faily simple, I'll work on it

bitmap: You're a champ. ❤️

\o/

he is :D

I guess that's it for tonight's meeting.

yyaay!

Thanks for your time everyone! :)

</BANG>

Woo!

thank you, freso :)

alastairp: Is the McGill dataset the one from the talk at MTG I went to?

Thanks Freso

bitmap: can you ping me when you've got the test server up and running?

I want ot .. eh.. test it :D

(also so I don't forget >__<)

CatQuest: ofcourse :)

bitmap: are MBS tests failing in general or is it PR specific?

I have no idea why tests are failing here https://github.com/metabrainz/musicbrainz-serve...

well, I haven't checked the tests closely, but you are assigning a variable to a package that doesn't exist :)

huh?

JSON::2::Role::Rating was removed recently, perhaps you had an older code checkout

yes I think

let me check

it also uses a different method for doing that now https://github.com/metabrainz/musicbrainz-serve...

yes, I saw

just calls serialize_ratings directly on the entities that needs it

okay

well , picard only needs it for recordings

but xml ws techincally includes it for artists as well

in that case it might be better to add back a package level variable

bitmap: whats LD?

Linked Data, as in JSON-LD?

right, no need to touch that

okay

was just wondering what it was

it's a fancy format we embed in our HTML for google, mainly

okay

so I need to add a new package variable to utils for ratings?

yeah, just 'our $force_ratings' or something, and check it in serialize_ratings

there are other things we output only for top level entity in json ws but for all linked sub entities in xml

okay

there's quite a few differences from the xml iirc, some might be good changes

though if picard breaks without something we'll need to add that

okay

it wont techinically break, it will just have to make call for each recording to fetch its ratings

which is more code and more requests

I figured it was easier making the changes on MBS

agreed

bitmap: how do I run tests locally?

to just run a particular test, prove -vl t/tests.t :: --tests 'WS::2::JSON::LookupRelease'

or 'WS::2::JSON' to run all JSON WS tests

hmm, do not have the test db setup

how do I init that?

./script/create_test_db.sh

thanks :)

wow, that's some pretty detailed test output :O

it's just TAP :P

still, pretty verbose

I know there's prolly a python lib out there to do this, but I would love to make a package with decorators that do just this

maybe something there https://testanything.org/producers.html#python

yeah, expected that :P