Notes from #MetaBrainz Meeting 2026-09-07

[aerozol]

Meeting start: [IRC Logs for MetaBrainz | MetaBrainz Chatlogs](Meeting start: IRC Logs for #metabrainz | MetaBrainz Chatlogs
Meeting wiki page: MetaBrainz Meeting - MusicBrainz Wiki

Agenda

• Reviews

Last meeting: 2026-08-31
Next meeting: 2026-09-14

@reosarevok: Welcome to another MetaBrainz Monday Meeting!

Reviews

@aerozol

[mailed in] A nice meeting with monkey to discuss some very interesting design/UX stuff.
Quite a lot of socials regarding LB troubles and Picard updates, plus some wiki updates, forum moderation, etc. The usual small fry.
Then drove 8 hours to play two release shows up north over the weekend, then did the 8 hours back. Tired.

@failure_san

[mailed in] [failure] I had to take a trip to the ER last week. Nothing much to report really, was out recovering.

@anshgoyal31

[ansh] Last week, I picked up a lot of old incomplete open PRs on ListenBrainz by GSoC Candidates
Picked up a few tickets to work on
I also found some time to complete my Old PR to revamp MusicBrainz Homepage
And started moving LB Services to Ansible - I’ve started with smaller stuff. After I’ve tested it - we’ll be moving all our services to Ansible which will make deployment easier.

@mr_monkey

[monkey] Last week I had some meetings with aerozol, zas, and the LB team, all to catch up on ongoing projects and The State of Things
Did a buinch of outreach to various third-party software/scrobblers that uses LB, to ask them nicely to add a user-agent string and respect rate limits. So far so good and everyone has been understanding
And a lot of PR review, like ansh taking up some older PRs to try and push them past the finish line.
In the works for example is Tidal support (both importing history and streaming on the LB website)
Did some ticket triage and general roadmap/priorities alignment for LB
Deployed a new version of BookBrainz with some database-related fixes
And looked into security tickets
And a bit of social media/email messaging to LB users.

@JadedBlueEyes

[Jade] This last week I’ve mostly been really tired. I managed to sneak in some MB related conversations and maintenance among other stuff, and I have some code review for reo queued up, but that’s about it

@Silona

[Silona Bonewald] Twenty CRM - let my brain get ahead of me?
Gusto setup finalish?
Two days Going over much on sponsorship proposal. Doing search and backup for the asks.
More of analytics and metrics - patterns?
Understand zas AI defenses
Getting a handle on Infra structure costs to help with the sponsorship asks
Prepping for sept 17th board meeting and docs
i have to say from what I have seen so far from zas defenses I am impressed AND also see why cloudflare wouldn’t work for us easily…
I have to resist the urge to really nerd out on it though. But I will be nagging EVERYONE about budget stuff for 2027.
I want to do two budgets - an as it budget and a growth budget.
so heads up!
so any ESP infra budget asks - send to me today!
i’m betting LB has some asks…

@bitmap

last week I was finishing up Web service performance improvements (2026 Q4) by mwiencek · Pull Request #3852 · metabrainz/musicbrainz-server · GitHub
after working on some more optimizations I’m seeing median WS response times drop by quite a bit locally, but don’t want to hype it up too much yet
also worked on adding a basic UI + editor column to the ‘noindex’ PR, updating a couple other PRs, doing code review, updating beta

@garvthakre

so i was on little break as my exams are also going on .
so nothing much from my side . but i think i will start working on some issues .

@Haris

First half of last week was quiz, later half was me being sick. Mood-tags research had to wait unfortunately.
ugh no work on my PRs either
not a fun week but ended it nicely with a movie : )

@reosarevok

Today I’m feeling fishy
More seriously, last week I did some actual non-mail stuff! Mostly converting the templates for autoeditor elections to the mb-mail-service style so we can hopefully automatically send them next time there’s an election
Dammit, that’s actually technically mail stuff again
I also did a bit of code review for bitmap’s WS improvements, that hopefully will be as good as they promise to be
I pestered lucifer to add a way to metabrainz.org so we can mark pre-revenue “use the data for a bit then come back if you make money” people better
And I dealt with a lot of new signups and the like

@ApeKattQuest_MonkeyPython

[MonkeyPython] still working on INST-1160
(alas)
will issue a PR for images Soon™
the hallway electrician are being slowpoking feckers so alas have not moved on there.

• INST-1160: improve Tambourim

@jasjeet

[jasje] Last week was tad bit looking into making login better + google play reviews
Also tad bit PR reviews
planning for summit

@yateen

Last week I didn’t contributed much , I was busy in college assignments and preparation for placements.
However, I modified one of my PRs for the project based on Ansh’s review.

@Gopal7387

[Gopal] I focused on my remaining work on the LB-1942 and LB-1651.
Especially for LB-1651, solved a sneaky text-overlap bug during live testing and finally got a better version out of it. Still under review, monkey ; please do check that out when you get a moment
For LB-1942, thanks for confirming the direction, monkey ; going with the better version. I’ll revert the old scroll-container work and raise a fresh PR for it this week.

• LB-1942: Order of albums on artist page should read left to right then down
• LB-1651: Add release/listen text options to art creator/stats art generator

@lioncat6

This last week I’ve been working on polishing the latest SAMBL update before release. Subvert is now fully functional and the seeder is avaliable for testing.

@zas

Yet another very busy week for me.
We released Picard 3 RC1. I addressed multiple issues in preparation for RC2, either issues I identified or reported by users.
I fixed several potential crashes, malformed profile files, broken plugin and registry data, and stopped OAuth secrets from leaking into debug logs.
I did a round of memory work: moving album cache to disk (PICARD-3409) and storing single-valued tags as plain strings.
I also polished the UI: async actions (PICARD-3410), a view refresh on color change (PICARD-3417), unified boolean env var parsing (PICARD-3414), and safer pygit2 timeouts (PICARD-3420).
I also reviewed a bunch of my teammates’ PRs.

• PICARD-3409: Reduce memory usage
• PICARD-3410: Show a busy cursor to acknowledge async lookup and cluster actions
• PICARD-3417: Item views and the metadata box do not refresh correctly when interface colors are changed and applied
• PICARD-3414: Unify boolean environment variable parsing and add PICARD_FORCE_FUSION
• PICARD-3420: Set proper pygit2 timeouts, reduce the risk of hang

We plan to release RC2 on Thursday.
I wrote a report aimed at the team and new ED regarding the current state of the infrastructure. I also put some thoughts on the future SOLR cluster and we came (thanks yvanzo) out with a plan (and a budget).
I started to address some of the issues noticed during my audit.
I set up Consul observability: I wired Consul’s Prometheus telemetry to telegraf, then tuned the metrics; stopped duplicating cluster-wide metrics across agents, split agent-local from server metrics, and curated the fields. I added ZFS monitoring, a Consul health-check input, and brought quest in as the 5th Consul server.
I improved the telegraf and consul Ansible roles: multi-group config merging, a post-restart raft-health gate, and check-mode fixes.
Also I worked on a uwsgi-to-HTTP migration across several repos: dual socket and health endpoint for picard-website, the new HTTP views in docker-server-configs, and auto protocol detection in the openresty-gateways templates. I also released serviceregistrator 0.8.1 with tag-aware multi-port naming to help in the process.
I documented everything regarding this for the rest of the team.
On bot-detector, I finished the CLI client (botctl), giving operators a simpler way to interact with bot-detector (mainly reo & me).
Plus some grafana dashboard & alerts improvements, system upgrades, and usual user support & system supervision.
I also had interesting video calls with yvanzo (about solr cluster), monkey (about everything) & aerozol (about linkbrainz).

@outsidecontext

zas said most already. After the rc1 release we had several issue reports, the new version clearly gets used. Overall very positive feedback
I did some PRs with bug fixes, including a crash and missing release info in the appstream metadata (used for Linux packaging). I also opened a PR to prepare updating the flatpak build for version 3.0.
Plan is to do rc2 on Thursday, and final 3.0 later this month.

@julian45

did some maintenance work this week:

• cleared a backlog of dependency upgrades this week for mb-mail-service (including adapting to a couple of breaking changes in dependencies)
• wrote up internal docs for the staff SSO system
• grafana updates for security
• patching here and there

@yvanzo

Since last meeting I drafted a patch to block deep searches in MB.
(See comments and attached statistics to the ticket MBS-14455.)
Also worked on tagging requests from MBS to MB Solr.
Had an MB dev meeting with bitmap and reosarevok , reviewed pull requests.
Last, reviewed SEC tickets for MBS and fixed the automation rule for alerts to be resolved.

• MBS-14455: Limit the depth of search to 500 results

@kellnerd

[after meeting] After my semi-finished move, I’m currently stuck without access to my computer and no internet at home.

absent: lucifer (1), Nirvan Jain (2), fettuccinae (2)