#metabrainz

/

      • Jigen joined the channel
      • LupinIII has quit
      • LupinIII joined the channel
      • Goemon joined the channel
      • Jigen has quit
      • LupinIII has quit
      • ApeKattQuest joined the channel
      • ApeKattQuest has quit
      • ApeKattQuest joined the channel
      • dseomn has quit
      • dseomn joined the channel
      • d4rkie has quit
      • d4rkie joined the channel
      • fettuccinae[m]
        <aerozol[m]> "@fettuccinae:matrix.org: last..." <- Hey, i pmed it to you few days ago
      • aerozol[m]
        Hey! Oops I forgot about the pm soz 😅
      • wileyfoxyx[m] joined the channel
      • wileyfoxyx[m]
        Got notifications from this channel for some reason, and was about to say @fettuccinae:matrix.org that Matrix-to-Discord mirror doesn't support PMs yet because I thought it was directed to me 😁
      • TOPIC: MetaBrainz Community and Development channel | MusicBrainz non-development: #musicbrainz | BookBrainz: #bookbrainz | Channel is logged and not empty as it is bridged to IRC; see https://musicbrainz.org/doc/ChatBrainz for details | Agenda (May 26): Reviews
      • Kladky joined the channel
      • reosarevok[m]
        I forgot to bring up yesterday that since next Monday is schema change day, we will NOT have a standard meeting! Next meeting is May 26! ansh, bitmap , julian45, lucifer, mayhem, monkey, yvanzo, zas, LupinIII (IRC), outsidecontext, jasje, Gautam_coder, Jade , kellnerd, aerozol, @fettuccinae:matrix.org
      • q3lont joined the channel
      • q3lont has quit
      • yvanzo: I'm off for a little bit but I'll be back in an hour at most I think - let's look at more search stuff? :)
      • Also, when around, please see https://github.com/metabrainz/mbsssss/pull/69 - I think we should merge that (and release mbsssss) for https://github.com/metabrainz/mb-solr/pull/63
      • There's of course a bunch of other open PRs there that could be part of such a release, but that depends on whether it makes sense and we have the time
      • SigHunter has quit
      • SigHunter joined the channel
      • aerozol: we need to schedule some schema change posts in mastodon and bluesky, do you have some time to help with that?
      • (between now and Friday would be ideal)
      • yvanzo[m]
        Hi reosarevok: I’m updating docs for releasing MB Solr w/ Solr 9. Feel free to test #50.
      • reosarevok[m]
        Can do
      • By the way, in the docs:
      • `time for host in $(ls ../docker-server-configs/scripts/nodes | sed 's/\.sh$//'); do echo; echo "VVV $host VVV"; sshc "$host" sudo su - root -c 'cd /root/docker-server-configs/ && git status --ignored && git fetch --all --prune && git pull --ff-only'; done`
      • sshc does not seem to be available in these servers - did this actually work last yer?
      • s//`/, s//`/, s/yer/year/
      • yvanzo[m]
        It is supposed to be run from your own machine. sshc is defined in docker-server-configs.
      • reosarevok[m]
        Also, I tried to do the " Update the MusicBrainz API Search examples in XML and JSON" bit for mine, but that part actually gets updated immediately in /doc pages because of how the transclusion works
      • So I guess we have to choose: do we do it in advance and it is wrong in the docs until release, or do we wait until release and do it after
      • yvanzo[m]
        in advance and add a disclaimer such as: this page is in advance
      • reosarevok[m]
        yvanzo[m]: Hmm. I was running it in my own machine, from my docker-server-configs folder 🤷‍♂️
      • yvanzo[m]
        So you have to define sshc in your shell.
      • Just copy the function to your ~/.bashrc
      • reosarevok[m]
        Ok, thanks, running nowç
      • Done
      • For the docs, do you mean you'd just add a banner at the top of https://wiki.musicbrainz.org/MusicBrainz_API/Se... that says "this includes the features to be released soon as part of the MusicBrainz 2025 Q2 schema change" and a link tot he blog?
      • If that seems good, I can do that for your PR and mine
      • Goemon
        lol, so this one doom/stoner metal norwegian band called slomosa has a song called "Horses" this one swedish doom/sludge metal band called domkraft has a song called "Horses Horses"
      • I'm eagerly a Finish stoner/sludge band with their song "Horses Horses Horses"
      • reosarevok[m]
        yvanzo: also I just realized we don't yet have AC mbids for non-search API results either, are we expecting to release that together with the search ones?
      • yvanzo[m]
        reosarevok: That will be in a second release, ideally this week too.
      • reosarevok[m]
        I did use the existing MBIDs for the artist credits, used a generated one for the mediums since we don't have that yet :) ç
      • (argh, I do hate where ç is in this keyboard)
      • derat[m] joined the channel
      • derat[m]
        reosarevok, bitmap: not urgent, but i'm curious about whether either of you have any immediate guesses about the source of the DMARC-failing messages in https://tickets.metabrainz.org/browse/MBS-14020
      • BrainzBot
        MBS-14020: Email messages are sometimes sent from users' domains and fail DMARC checks
      • monkey[m]
        DMARC is on DWALL — Virginia Woolf
      • reosarevok[m]
        Jade: any idea if that might be related to the new service? ^
      • Jade[m]
        It won't be, the new service hands the emails off to an actual mail server to deliver them, that should handle DMARC, etc
      • mayhem[m]
        @fettuccinae:matrix.org: ping. lets catch up and chat about your project. let me know when you're around.
      • Jade[m]
        Jade[m]: I'd have to check which software is specifically but it'll be the same one handling the old emails
      • petitminion joined the channel
      • reosarevok[m]
        Thanks
      • derat: so, no, I don't know - bitmap might
      • But we've had weird issues with emails for ages (user reports' reply to has been broken for years for example) so... not sure
      • derat[m]
        i think that the issue here is that the From: header has the user's address instead of a metabrainz.org address, so there's probably nothing the outgoing mail server can do at that point
      • i could try just setting my domain's DMARC policy to reject, but presumably the bounces will just go to /dev/null somewhere, so i doubt it'll help with figuring out which messages these are
      • reosarevok[m]
        I would assume those are then user-to-user messages with "show my email address"
      • fettuccinae[m]
        <mayhem[m]> "@fettuccinae:matrix.org: ping..." <- pong, im here.
      • mayhem[m]
        hey
      • fettuccinae[m]
        hello
      • mayhem[m]
      • Jade[m]
        <Jade[m]> "It won't be, the new service..." <- The from address set should always end up being the metabrainz email, it's set based on what the API call gives it https://github.com/metabrainz/mb-mail-service/b... which comes from https://github.com/metabrainz/musicbrainz-serve..., and from what I can see the only way the from
      • address would be wring is if somehow https://github.com/metabrainz/musicbrainz-serve... stopped anonymising properly
      • mayhem[m]
        in the original spec we had two more fields in the notification table. template_id and template_params.
      • lucifer[m]
        reosarevok: also for edit notes i think
      • fettuccinae[m]
        mayhem[m]: yes, if i understand it correctly, they are used for mb-mail. i thought to add those fields later in additional integration when integrating mb-mail services
      • derat[m]
        reosarevok[m]: that seems like a decent guess, but i think that i never check "show my email address", so i'm not sure why those would be sent
      • mayhem[m]
        lets add those now -- eventually everyone should be using the template system
      • Jade[m]
        Jade[m]: Basically I can't see a way it would ever get set to anything other than `MusicBrainz Server <norely@musicbrainz.org>` or `username <norely@musicbrainz.org>`
      • fettuccinae[m]
        mayhem[m]: okay, ill add it
      • Jade[m]
        Would be helpful to see the dmarc reports & the emails themselves triggering it, but I've got an exam in 40 mins 😅
      • <derat[m]> "i could try just setting my..." <- Does it not have a quarantine option?
      • Jade[m]: Also sender is flat out always the noreply address 🤷
      • derat[m]
        Jade: it does, but that probably won't help either since the messages would just end up in some gmail user's spam filter :-(
      • i'd love to see the emails triggering it too! re DMARC reports, i pasted the JSON form that the postmark API gives me in the ticket
      • Jade[m]
        I'll have a deeler look after my exam tomorrow morning
      • derat[m]
        thanks! (just to be clear, i don't think there's anything to suggest that these are coming from the new mail service. i only started looking at DMARC reports recently, so this could've been happening for a long time. also possible that the messages are sent by e.g. jira or discourse, although the messages that i myself receive from those services look okay)
      • fettuccinae[m]
        mayhem: ive added template_id and params. (i forgot user_email field in user table yesterday, so, added that too).
      • Jade[m]
        derat[m]: Yeah, until we can actually get the emails in question it's hard to say anything for sure. Email deliverability stuff is never fun
      • _BrainzGit
        [bookbrainz-site] 14MonkeyDo opened pull request #1169 (03master…update-dompurify): chore: Update DOMPurify https://github.com/metabrainz/bookbrainz-site/p...
      • [bookbrainz-site] 14MonkeyDo opened pull request #1170 (03master…update-superagent): chore: update Superagent https://github.com/metabrainz/bookbrainz-site/p...
      • julian45[m]
        <derat[m]> "i'd love to see the emails..." <- ah, DMARC. minor bane of my existence at $dayjob
      • let me see if i can poke at google workspace's logs
      • derat[m]
        julian45: thanks! per the report, it looks like there were 4 messages with `erat.org` in the `From:` header sent from google's servers to google-hosted users between `2025-05-05T00:00:00Z` and `2025-05-05T23:59:59Z`
      • julian45[m]
        any chance you have more logs beyond those you provided in MBS-14020?
      • BrainzBot
        MBS-14020: Email messages are sometimes sent from users' domains and fail DMARC checks https://tickets.metabrainz.org/browse/MBS-14020
      • julian45[m]
        something like a message id, to address, etc. would be helpful
      • (feel free to DM me for stuff that shouldn't be shared in a public, logged channel like this one)
      • derat[m]
        i fear that that's all you can get from DMARC reports, unfortunately. by design, i think that they exclude details about the individual messages; i believe they're just bucketed by something like (day, sender ip, spf/dkim details, result)
      • bbiab
      • _BrainzGit
        [bookbrainz-data-js] 14MonkeyDo opened pull request #325 (03master…upgrade-mocha): chore: Upgrade mocha to v11 https://github.com/metabrainz/bookbrainz-data-j...
      • julian45[m]
        alas, google admin console's email log search is kind of ass. if i search for emails sent between 5/4 and 5/5 (in my timezone since that's what the console allows searching time ranges with) with sender email including erat.org, it only shows the weekly meeting you mailed to MeB ~9:20am ET on 5/5
      • reosarevok[m]
        yvanzo: I tried running your branches but it doesn't build indexer, probably because either a) I'm doing something wrong and dumb or b) I have it all set up for py3 but your branch is still py2 :)
      • Maxr1998_ has quit
      • Maxr1998 joined the channel
      • derat[m]
        julian45: ah well, thanks for checking. based on the infrequency of these, i suspect that they're sent by some feature or service that i only occasionally use. no idea how to figure out which one, though :-/
      • julian45[m]
        <derat[m]> "julian45: ah well, thanks for..." <- np! if each of us had deeper access to email servers than we did, it might be possible to drill down deeper, but alas
      • mayhem zas : mind if i stand up a new tiny hetzner VM w/i the existing sso "project"? idea is to run a test service or two there to help put authentik through its paces integration-wise
      • * put authentik (the tool powering auth.meb.o) through its
      • mayhem[m]
        no problem.
      • how long do you need it for?
      • julian45[m]
        matter of days or weeks, most likely
      • mayhem[m] will put that in the calendar before creating a VM.
      • mayhem[m]
        give me a deadline please. at that point I will check in with you, but we have a nasty habit of forgetting these machines.
      • and what specs? what amd? arm?
      • julian45[m]
        mayhem[m]: fair point. call it 6/30 at absolute latest (i will be doing some moving between apartments later this month and may get held up on this front)
      • mayhem[m]
        lets sayt 6/15 and I will check in with you.
      • julian45[m]
        mayhem[m]: i was thinking a hetzner CAX11 (cheapest arm, EUR3.29/mo) should do the trick; if that was not beefy enough for some reason, i'd go to a CAX21 (EUR5.99/mo)
      • mayhem[m]: works for me
      • mayhem[m]
        ohh, I dunno that. that price seems a bit excessive. 🤣
      • shall we just go for CAX21?
      • julian45[m]
        works for me
      • i have enough access to hetzner now to spin it up -- ok if i proceed?
      • mayhem[m]
        do you care where? Helskinki seems the default/cheapest.
      • hells-kinky! that sounds interesting.
      • julian45[m]
        i was gonna do nuremberg or falkenstein, in my experience i get measurably lower latency from those datacenters than helsinki here in the midwest US
      • mayhem[m]
        what would you like to name this machine?
      • julian45[m]
        authtestbox, perhaps?
      • julian45[m] shrugs
      • mayhem[m]
        do you need a ipv6 IP?
      • ah, no that is included.
      • julian45[m]
        right, and it gets a few tens of cents cheaper per month if you cut out the ipv4 -- i happen to have ipv6 here at my apartment, but not always around here
      • mayhem[m]
        no need for that.
      • pm'ed you the IP. let me know if you can't get in.
      • julian45[m]
        got in, thx!