Notes from #MetaBrainz Meeting on 2016-05-16

[Freso]

MetaBrainz Meeting 2016-05-16

Meeting start: IRC Logs for #metabrainz | MetaBrainz Chatlogs

Agenda

• Reviews
• Schema change announcement
• Developer priorities
• Picard 503

Reviews

@Freso

IRC Logs for #metabrainz | MetaBrainz Chatlogs

@Freso

Over the last week I have…
Spent some time reaching out to other communities (Mixxx, Quod Libet, VLC), which has resulted in a couple of non-profit sign-ups as well as MeB being listed on VLC’s website: Our partners - VideoLAN
Been around and about.
And earlier today I finally got around to finishing up the HTTPS announcement blog post as well as helped ruaok out with a reviewing some other blog posts.
[…]
I also added the MetaBrainz blog to https://planetgsoc.github.io/
Students with a blog may want to add themselves there as well.

@chrisskye

IRC Logs for #metabrainz | MetaBrainz Chatlogs

@chrisskye

I’ve been suffering from a chest infection more than a week, so a bit sluggish. That said, managed to do a lot of correspondence. Lots of new sign-ups. A few are paying supporters, some non-profit and community orgs and numerous start-ups. Rob and I continue with the usual run-around trying to get service hook-ups to the new office and dealing with the new Spanish bank account. I’m off to Calgary, Canada tomorrow. Will be working from there. So closer to Pacific Coast time, than European time.
[…]
I received an email from SiriusFM (Sirius XM, red.) just now…asking a lot of questions…but basically wanting to sign an agreement for Unicorn Tier.

@reosarevok

IRC Logs for #metabrainz | MetaBrainz Chatlogs

• dealt with support, including delegating some stuff to chrisskye (yay!)
• got tired of not being able to set user emails when they didn’t get our verification, so I wrote that code
• it’s in beta now, and I already had to use it once, so I’d say it was a success
• spent the second half of the week worrying about opening the museum I volunteer / intern at, which means 2-in-1 release groups and other boxsets (STYLE-330, STYLE-331, STYLE-335) still needs me to decide stuff which gah I’m sorry I’ll get to it
• (it’s not very straightforward at all so I need to consider it carefully and I just haven’t had enough peace and quiet to think about it)

@bitmap

IRC Logs for #metabrainz | MetaBrainz Chatlogs

• did schema change stuff

@Gentlecat

IRC Logs for #metabrainz | MetaBrainz Chatlogs

• added new customers to meb.org page
• reviewed GSoC work

@zas

IRC Logs for #metabrainz | MetaBrainz Chatlogs

• hacking on new rate limiter
• search servers

@Rob

IRC Logs for #metabrainz | MetaBrainz Chatlogs

@Rob

apparently quickbooks never thought of having an invoice in USD and payment in EUR.
that is going to be a real headache for me.
so, I spent way too much time on that. then zas appeared in town and we’ve been catching up.
we threw a really awesome house party, but I guess that is OT.
more seriously we’ve been putting serious amount of time into the search server.
we’ve discovered two particular problems. we’ve fixed them too, but we’re still possibly battling some deployment issues.
part of this is motivated by the fact that we need to move to NewHost soon.
and zas and I are looking at ways to try and reduce the time we spend in order to get to NewHost.
so, we’ve been researching NewHost and making good progress on that.
We’ve got figures for OVH and Hetzer and awaiting others.
not going to the full SOLR route, but keeping the search servers might shorten the time to NewHost by a few weeks.
but, there is hope for fixing the current search servers.
now we’re focusing on the schema change release, which we had to postpone for a week… because reasons.
I don’t think that any of us felt really comfortable with where we were.
so, rather than push and make life even more miserable for the jet-lagged bitmap, we decided to postpone it a week and make sure we get it right.
the plan now is to go to PG 9.5 tomorrow morning. and then the schema change release in a week.

@LordSputnik

IRC Logs for #metabrainz | MetaBrainz Chatlogs

@LordSputnik

This week I did some planning for BB OAuth, and I’ve mapped out the flow that users should go through to register, log in and link existing accounts to MeB accounts.
I’m quite happy with all that now and if I get time this week it’ll be on our beta server by next weekend
Leftmost and I also had a long discussion with QuoraUK to plan various aspects of his GSoC project. We’ve decided on the principles of the system, which QuoraUK can talk a bit more about later in his review.
We started putting together a list of the achievement types we’d like to see: Achievements - Google Docs
And we’re aiming to discuss a solid plan/set of milestones for implementing the system at next week’s BB meeting.

@chirlu

IRC Logs for #metabrainz | MetaBrainz Chatlogs

• did a number of schema change tickets
• also prepared auto-pestering for banner message editors so that they will take down banner messages sooner

@rahulr

IRC Logs for #metabrainz | MetaBrainz Chatlogs

Mine end semesters ended last week. That’s quite a relief for me.
I wrote a small blog post about gsoc selection.
Resumed working on picard-258, that is about visual feedback on coverart changes.
And looking forward to discuss some project implementation details with bitmap later on this week.

@Quora

IRC Logs for #metabrainz | MetaBrainz Chatlogs

@Quora

wrote a blog post about the design for the gamification system, which we discussed in the BB meeting, got some good ideas for achievements in the doc above, Got a package from Google today and will work on project planning ~thursday once my exams for the week are over
should get a new post up friday detailing ready for discussion saturday

@armalcolite

IRC Logs for #metabrainz | MetaBrainz Chatlogs

@armalcolite

This week i was focused on doing task which ruaok had assigned
Created a new PR: https://github.com/metabrainz/listenbrainz-server/pull/80
It is an attempt to store 30-90 days of listens in postgres for analytics generation.
I

• Added support for listenstore.conf in listenstore. (It was mentioned as TODO in the source code)
• Added support for storing listens in postgres, includes updating schema and code modifications.
• Added Support for scheduling jobs using apscheduler. Job for clearing postgres DB every 24 hours to maintain 90 days listens.
• A new script to push listens to postgres instead of cassandra, “listenstore-kafka-to-postgres-writer.py” in addition to older one.
  The listens can be pushed to any of them by just switching a single line. (Provided the schema is correct)
• Reading listens for display on site from postgres instead of cassandra. (This also can be switched by changing a single line)

@hellska

IRC Logs for #metabrainz | MetaBrainz Chatlogs

@hellska

last week worked most of the time on the project!
rewriting the proposal after the meeting and answering the comments
digged into the AB database schema to understand where to make changes
followed the reviews on dataset API and dataset validation

@kartikgupta0909

IRC Logs for #metabrainz | MetaBrainz Chatlogs

@kartikgupta0909

Finalised some of the details about the project with alastairp, which will help me get started about the project
We have decided on making a PPA for the offline dataset training tool
and will make that PPA this week itself so that next week I can devote time to write the installation shell script.
and also I started working on my first track as a music producer

@gcetusic

IRC Logs for #metabrainz | MetaBrainz Chatlogs

@gcetusic

Hi! I’ve been working on creating tickets for AB for the first few weeks of GSOC: #125, #126 and #127 for now. I’m focusing on designing the schema which I hope I’ll kick off this week but I’m still waiting for Felipe to answer the last email alastairp sent about me getting access to his storage bucket so I can play with BigQuery. Otherwise, I’ll do it on my own account but I’ll have to load the summary data myself to cloud storage. Other than that, I’ve cloned the repo of the acousticbrainz server, set up my virtualenv and poked around a bit but to get a feel for the code because I’ll have to change it at some point when uploading data to bigquery. Last week I completely forgot about the meeting! [OT] Yesterday I arrived in Dublin! Wooo!

Schema change announcement

IRC Logs for #metabrainz | MetaBrainz Chatlogs

@Freso

@Rob already pretty much said it in his review, but we’ve pushed the schema change a week to try and be sure we have enough time to check that nothing breaks critically.
Important: Schema change delayed to May 23 – MetaBrainz Blog for more details.

Developer priorities

IRC Logs for #metabrainz | MetaBrainz Chatlogs

Announcement from @Rob that once the schema change release is out, the development priorities will get shuffled around a bit:

1. Hotfix critical bugs as they appear.
2. Coordinate with @Zas to identify and fix issues that prevent us from moving to NewHost.

@Rob

All hands on deck until we can move to NewHost.
I am taking the busywork, paperwork side of things.
@zas is the chief architect. @bitmap is the lead engineer and @Gentlecat will be a senior level minion.
the primary goal is: move to NewHost in an MVP (minimally viable product) fashion.
if a task needs doing and it is either: 2 days or 3 weeks, the answer will nearly always be 2 days.

Picard 503

IRC Logs for #metabrainz | MetaBrainz Chatlogs

Basically we’re seeing a lot of complaints about Picard users being pretty much unable to do anything due to getting 503’s from MusicBrainz, so @reosarevok asked if we could tweak the new rate limiter a bit. Short answer: yes, @Rob and @zas will continue to work on and tweak the rate limiter so hopefully Picard users should get a better experience shortly.