#metabrainz

alastairp: But isn't that exactly what your comment towards the end addresses? Adding a `use_https` parameter defaulting to `False`?

If it was just s/http/https/ it wouldn't be big enough for a GCI task. :)

[CAA-indexer] mwiencek opened pull request #14: Refresh connections (master...refresh-connections) https://git.io/vNTqD

bitmap: any chance you still have the link for the tiny dev dump?

Hey I'm from GCI, I'm adding releases to musicbrainz data

Now I'm unable to find more CDs at my home

What should I do

Rahul_india: #musicbrainz is probably a better place to ask for help with MusicBrainz editing. :)

Rahul_india: well, you can try some other tasks out

Rahul_india: the 78 rpm task is a good one

yvanzo: I think I've got my local MusicBrainz instance running again - but the last french test is failing (the Troie one) even with keepuppercase. is it reproducable on your end?

yvanzo: It doesn't seem to be related to my changes, since the test is failing becuase "Troie" is changing to "troie"

if I change the test to expect "troie", it passes

yvanzo: what it looks like is that French mode, for some reason, doesn't respect keepuppercase

which is very odd

yvanzo: the Jenkins build is also failing for what looks like the same reason

confirmed not related to my change because test still fails without it

naiveai: I don't have it anymore, but I'm generating a new one now

bitmap: thanks!

naiveai: replied to your task, I'll review the PR later today but it looks pretty good so far :D

Leo_Verto: :D

between you and annebelle, I'm really running out of BrainzBot tasks. I guess I'll have to turn some more of the tickets into tasks then :P

Leo_Verto: actually, I think this is my first BrainzBot task - all credit goes to annebelle

Leo_Verto: done!

[critiquebrainz] IsaacZanoria opened pull request #170: CB-263: Update publication date of draft reviews on CritiqueBrainz when they get published (master...CB-263) https://git.io/vNT8V

I'm trying to test the renderRelationship function, which returns a string of rendered HTML. Is it worth bringing in a dependency like chai-html for this?

Leo_Verto: for the document & test task, are we allowed to modify the function in question to pass the tests?

Leo_Verto: and do we have to check things that flow type annotations can? in my case, renderRelationship(relationship) requires that relationship have `source` and `target` properties, but that can be checked using a flow type annotation

hello i'm from GCI

rahul_india: yes, no need to introduce yourself again :)

i've a CD which has tracks in .cda format. picard is unable to detect any files on the disc

CD lookup is opening but not showing any files in it.

This might help rahul_india

i downloaded a cd ripper

its working well

thanks :)

no problem

zas: what's x-accel-redirect?

samj1912: it is a request type where the backend sends the whole response of a request is sent to nginx to return to the caller.

that sentence flew over my head

without it, if you have slower requestor, say on a mobile connection, the backends needs to wait until nginx has finished sending all the contract before the backend can handle a new request.

which ties up precious back-ends.

ah

but by handing the whole request to nginx and then going off to handle the next request, the back ends (expensive) are freed up quickly and the cheap front ends kick around until the request is done or times out.

not sure how to handle this https://github.com/metabrainz/musicbrainz-serve...

let me try, if I understood it correctly

I am assuming it basically feeds https://musicbrainz.org/ws/2/artist?query=queen... to search.mb.org?

Freso: there's a slight error in https://goo.gl/ughsb1, the branch name is "viewport-meta-YOUR_GCI_USERNAME" rather than "client-function-YOUR_GCI_USERNAME"

sorta.

its hands it to nginx

the web service figures out the new URL for this, a search url, then hands it back to nginx, which then issues that request and returns it to the user. transparently.

makes sense

yvanzo: bitmap any idea why testing PRs for mbs is taking so much time?

ruaok: with SOLR-16 fixed, I guess no output incompatibilities should be left

SOLR-16: relationship target gets rendered in different way for area results https://tickets.metabrainz.org/browse/SOLR-16

matching craptastic sort order now?

ah, not that, the schema output

h

ah

for that, I will have to rewrite the analyzers and boosters

which might take the majority of this month considering I will have to look into lucene (which I hane very less experience with)

should most probably be lesser time

in particular look at the existing analyzers, especially the artist one.

its... insane.

https://github.com/metabrainz/search-server/blo... ?

wtf lol

we are specifically booting bach? XD