Notes from #MetaBrainz Meeting 2017-04-24

[Freso]

MetaBrainz Meeting 2017-04-24

Meeting start: Making sure you're not a bot!

Agenda

• Reviews
• Multi‐level Picard Work tags and classical music (PICARD-1043)
• MineoBot

Reviews

@bitmap

Making sure you're not a bot!

@bitmap

I worked on this script for dumping the db as json
to speed up the script I worked on MBS-6459 as part of it
but idk if we’re going to allow that publicly
also, I spent a lot of time profiling the ws with NYTProf and made a bunch of improvements, which helped a lot
was working on setting up a slave db on williams at the end of the week so we can make a full dump eventually

@Rob

Making sure you're not a bot!

@Rob

more of the same for me. a mix of biz dev/invoicing with quesito and such.
still playing go-between with a prospective supporter who has some custom requirements.
which is what bitmap is helping with in the form on json data dumps.
it will be nice to be able to point people to a non-pg database dump.
not as good as shipping our data in excel format, but a lot more approachable
I spent a bit of time trying to finish the rabbitmq LB PR up and finally got that up today.

@Freso

Making sure you're not a bot!

@Freso

In the last week I’ve dealt with a reported editor and kept up with the forums and IRC. Including trying to make a broad ban to hopefully keep the “allah is great” guy/bot out (for a little while anyway).
And I finally got around to getting my Spotify2MusicBrainz.py script to workable condition: Freso / spotify2musicbrainz · GitLab – feel free to play around with it.
Right now I’m (hopefully) on a plane on my way home to Barcelona. Not gonna miss the weather up here…

@reosarevok

Making sure you're not a bot!

@reosarevok

I was mostly doing support
And fighting my project that I need to finish to graduate school
I looked into the events bit, but decided that I want to wait with that to hopefully have place credits
Also, today I sat on the monthly meeting of Wikimedia Estonia, few interesting things going on there, nothing of direct interest to us yet, but​that might change

@Gentlecat

Making sure you're not a bot!

@Gentlecat

I finished most of the work for EUR donations support
tried to deploy it today, but there were some issues
it’s hard to test because of the way the whole thing works
I’ll need to set up a separate instance to be able to test paypal

@yvanzo

Making sure you're not a bot!

@yvanzo

almost the same as previous, weather apart.

@Quesito

Making sure you're not a bot!

@Quesito

email follow ups for accounting upgrade organization and confirmations of payments coming in. Which lead into some supporter reach outs and discussions.
Got some feedback from Ruaok about Verbiage changes to MeB–which I need to review with CallerNo6 and started in on ORG-14
This past weekend I entered my first pieces of data to MB. It’s pretty intricate, although with the guidance of gcilou’s soothing voice from the video tutorial and a billion tabs open–it was a piece of cake!
In all seriousness, it was quite enjoyable, and also I got a bit of an addicting feeling…and I can’t wait to input more!

@samj1912

Making sure you're not a bot!

@samj1912

I spent most of the week busy with college end sem assignments. Apart from that did code reviews for picard and PW. That’s about it. Now gearing up for theory exams. And will get back to coding picard stuff tomorrow. I was thinking of applying to fedora flock 2017 for a talk proposal about MB and Picard. I’ll need help with that if possible
and would like to know if MeB already has any plans about flock 2017?

@iliekcomputers

Making sure you're not a bot!

@iliekcomputers

I have mostly been busy catching up with coursework and lab assignments this week, so relatively little MeB work.
But, I sent a PR fixing LB-152 (it still needs some work before it is ready for merge)
I also got ruaok’s rabbitmq PR working on my setup, that’s it. fin.

@ferbncode

Making sure you're not a bot!

@ferbncode

this week I pretty much did the same stuff i did last week and sent PRs for the last two db models in CB.
hat’s all for my week :P. End Semester exams hitting next week

@Leo_Verto

Making sure you're not a bot!

@Leo_Verto

I arrived in England yesterday so I spent pretty much the entire week packing
I’m not sure how much work I’ll be able to get done from here

@Sophist

Making sure you're not a bot!

@Sophist

Addressing review comments on one Picard PR and two plugins.
And lots of discussions about work tags.

@caroline-g

Making sure you're not a bot!

@caroline-g

I’m back and ready to rock and roll

Multi‐level Picard Work tags and classical music (PICARD-1043)

Making sure you're not a bot!

@Sophist

I have asked to have this discussion because I think we need to make a strategic decision on how Picard tags Works in order to stay compatible and comparable with other tools and prevent Picard falling behind.
The issue has been prompted by recently introduced iTunes support for Classical Music and multiple Work tags i.e. what iTunes calls Work and Movement and what I am calling Main Work and Track Work. However this is NOT about providing compatibility with iTunes because the iTunes functionality is only a partial solution at best, and its mapping to tags is so poorly conceived. Rather the iTunes functionality has prompted other tools like Jaikoz/SongKong and Muse to try to implement multi-level Work tagging and really this is something that I think should be native in Picard.
There has been extensive discussion about this in PICARD-1043, but other related tickets are: PICARD-1049, PICARD-1050, PICARD-1068, MBS-9304. There was also prior discussion also in How to explain Works. However before everyone rushes off to read these, I will try to sum up the issue as I see it:
As we all know, the MusicBrainz concept of “work” is quite generic i.e. includes a hierarchy of works and part works as well as stand-alone works.
Pop tracks generally have only a Track Work.
Classical music often has Track Works (i.e. a part of a Movement), Movements and a Main Work, though sometimes (e.g. in Operas) Movement might be split into e.g. Acts and Scenes and sometimes there is an Overarching Work (i.e. Wagner’s Ring Cycle, or Vivaldi’s Four Seasons).
Musical Theatre often has Track Works and a Main Work.
There are probably other uses that I am not aware of.
To complicate things a bit, MB Works can have unlimited levels of “work is part of work” relationships, and whilst we can navigate our way to the top level relationship, there are no semantic attributes to tell us whether it is the Main Work / Overarching Work (if indeed these are clear cut in all cases) nor what the intermediate levels might represent etc.
Picard has for some years used the Track Work as the entire “Work” which is both somewhat simplistic and also imprecise terminology. iTunes now supports in its UI both the Main Work and Movement but its mapping to tags is poorly thought through. Jaikoz/SongKong does something similar, and wants to extend this. Muso have submitted a work-parts plugin to provide support too. I believe that Picard needs to provide better support on this, and if we agree to move forward it will likely mean some MBS development as well as Picard.

It was agreed that this was a bit much to digest and discuss at the present meeting, so it was suggested that @Sophist writes up a document of some sort and allow people some time to digest it before bringing it up again at a future meeting.

MineoBot

Making sure you're not a bot!

@Rob created a VM to host MineoBot but didn’t know at the time that the bot requires a full replicated MusicBrainz database to work. @Rob suggested running the bot on the instance hosting the FreeDB gateway instead and asked if anyone was using the instance MineoBot was currently assigned to. As no one spoke up, @Rob said he’s save a snapshot of the instance and then remove it.