#metabrainz

one traceback and then iliekcomputers dups. let's hope that is all there is.

iliekcomputers: https://tickets.metabrainz.org/browse/LB-2?jql=...

LB-2: Getting logged in as the wrong user

Freso: I suspect that this has to do with us blowing away the DB and your user id changing.

Possibly.

have you logged out and back in?

an old session might kick around, but a logout should fix that.

No. I didn't want to do anything until I had gotten some kind of heads up on what I should do in case further investigation was wanted. I'll try doing that.

maybe related: I looked at https://beta.listenbrainz.org/user/CatCat before i logged in and it was "page not found"

Logged in as Freso now, so doesn't seem to be LB-2 reappearing.

LB-2: Getting logged in as the wrong user https://tickets.metabrainz.org/browse/LB-2

1M listens!

:O

phew ok.

is there any live stats page we can look at? :O

\o/

ruaok: did you get any issue with listenbrainz-web docker logs since the upgrade ?

not yet.

I just tailed logs a pile of times, no crashes yet.

"As part of a Google Summer of Code project, we're going to add statistics and graphs for users, artists and lots more. This will be released a the end of summer 2017. " wehey!

this is also what iliekcomputers is working on?

that is what he is working on.

!m iliekcomputers

You're doing good work, iliekcomputers!

A fix to the user login thing should be to change the secret key each time we wipe the database

Then the logincookie will be invalid

+1 to that

well as long as it doesnt name it impossible ot login

but since datawhipes shoudn't be happenig in the future after beta-launch..

but. re the scrobbler for ppc...

Yeah, only something to do before beta launh

Lunch

it's hella more stable than the intel mac :/

I think they went wrong with switching to intel, honestly

I can pay like idk 200 nok?

anyone want some pocket money and help a nice cat out.

also then lb will *have* a ppc scrobbler (listener?) and the more of those the better I say

also: I cna wait. gods I ca wait, a year, 2, 4

likly every other machine will get a listener before ppc

we should invent a term that isn't scrobble

Last.fm: 39227 / beta.LB: 39227

!m iliekcomputers

You're doing good work, iliekcomputers!

!m ruaok

You're doing good work, ruaok!

you cna use "lytte" it's norwegian/danish? for "listen" amd yo ucna verb it nicely

a "lytter" is something that listens

Really looking forward to having LB hit public beta. :)

ahh. I'm not even half way done :/

I am very anticipational :O

Freso: mind doing a last.fm import again to see if there are any dups this time? It would take a few seconds or so

iliekcomputers: Again? On top of the one I just did?

Yeah

Ideally there shouldnt be any duplicayes

iliekcomputers: does that apply to me too? (when done)

MB has a surprisingly old userbase, right now the median seems to be somewhere in the mid 30s

Well, I just submitted a scrobble from my phone, verified that Last.FM received. Did another Last.FM import on b.LB, and b.LB only went 1 listen up.

iliekcomputers: ^

CatQuest: Nah, its okay, thanks :)

Leo_Verto: well we where younger whne we started. also what you mena *old* 30 isn't old come on.

:P

Freso: so, no duplicates, nice

:P

compared to a lot of other websites, I'd guess it is

those websites are what? social craps?

I love the person who responsed "I Sing well in the shower" to the questions about being a musician <3

lol

Considering the "archivist" nature of the side, I wouldn't be surprised to have more people 40+ than 30-.

*site

CatQuest: quite a lot of people in their thirties and fourties

you have to take in account mb is over 15 years *old*

also: survey taking bias

someone even answered with 81, that'd be pretty impressive if true

imagine all the data this person could add and has a perspective on :O

maybe 65 year old concerts :O

also so far the country the most people live in is Germany :P

survey time bias tho

but that might just be because of timezones

wait until it's morning in like. idk australia

indeed

how many responses so far?

131

also are answers not linked to (optionally if added) usernames?

CatQuest: oh you will :D

\o/

CatQuest: I'll seperate the demographics data and handle that seperately

ruaok: I can't reproduce the dup problem on my local setup

that's pretty good already :O

but having the usernames attached to the rest of the survey can be quite useful if we want to get back to certain people, e.g. to ask for clarification

Leo_Verto: well i hope yo ualso show everything stat as well as split up by demo stats

oh yeah, the demo stats are totally gonna be their own blog post :D

(if Freso lets me do that :P)

Leo_Verto: what I meant with that is that ya'll know what I asnwered to all the questions :D

i can't imagine why not

demo stats?

Leo_Verto: Of course.

demographic data, like where our users are from, how old they are, what languages they speak

ah!

which imho is the best part of the survey

i'd like if severla people write simmilar things in freform that is a stat too.. but I suppose that might be pushing it and be more work

also US answers seem to have overtaken Germany now, but they're split into "US", "USA", "United States" and all of those in lowercase

but I'm very glad I managed to write out my ideas for the votting thing a lot

you'll have to merge them :D

whooo the page sending is halfway now

I'm already seeing certain trends show up for example in the dislike question, quite a few people feel like they are just editing in isolation without coming across other editors too often

+100 to that

thats exactly the tig i think too

I miss having edits show up on the top of entites SO much. sicne you relaly felt other peopel were editing the database..

random no voting was an issue way back when we didn't have the coc, new users could vote from day one, much less clarity about what things were, after first no vote things didn't have a grace point

etc..

all these have been put into place to prevent random user who has no clue to vote down edits before they en knew what edits were

I know I for one would vote hella alot more if those where there

hva skjer if I start a second import in a separate tab before the first is over?

nothing bad, i'm expecting

try it :P

hn...

maybe someone will fewer listens can try that ;)

hah

hh. https://twitter.com/AcousticBrainz is in barcelona

aww she already follows me <3

someone is in love =3

what, no

for acousticbrainz, is the software (+training data) for calculating the highlevel features openly available? i would like to throw it at a large collection of music that isn't well-covered by musicbrainz and i'd like to run this offline too. (mainly interested in the mood features)

minus: yes, it's all available

unfortunately, might be difficult to find and set up

heh ok, i might try myself at hunting it down and getting it running then some time

https://github.com/metabrainz/acousticbrainz-se... is where the high-level extractor lives

but I can't find a link to the models that are used on ab.org right now

they are hosted on the university website somewhere

ah, here http://essentia.upf.edu/documentation/streaming...

cool, thanks

moment of truth

🎆\o/🎇

ruaok, iliekcomputers ^^

:)