#metabrainz

yvanzo: that for sure :)

Also, edit.display_data.${edit.display_data.entity_type} - how do I do this in React? :)

("nest" two variables like that)

reosarevok: IIRC, bitmap has been working on diff and other stuff might still be necessary to convert edit display (?)

iliekcomputers: :3.6.7 is better, but doesn't reflect updates to our base images

hello!!! I was setting up the local server for acoustic brainz and while running this command ./develop.sh run --rm --user `id -u`:`id -g` -e HOME=/tmp webserver npm install ..my terminal says "Run a one-off command on a service."..How do I proceed?

reosarevok: By the way, do you have a Solr instance setup along with your MB server?

the idea is that I shouldn't be able to break listenbrainz by pushing a bad python base image

yvanzo: yeah, he said I shouldn't touch stuff with diffs yet :)

And no, I don't

Don't have a local search server at all

prajjwal: mmmm, that sounds like the output from `docker-compose run`

alastairp: yeah, i guess that makes sense.

prajjwal: it's possible that your terminal pasted it in two lines for some reason, causing the full command to not run

reosarevok: How does that work? https://tickets.metabrainz.org/browse/MBS-8978

MBS-8978: Schema does not include id attribute for <packaging> element of a release

How does what work?

Cause solr doesn't have a clue about packaging :/

This https://github.com/metabrainz/mmd-schema/pull/18

iliekcomputers : As discussed earlier, I have written a script to push statistics in RabbitMQ and then eventually in the db.

I don't know enough about solr to answer this :/

alastarirp: So instead of copying shall i write it?

Maybe bitmap does?

It is just for testing purpose as of now.

To see how to handle data after being processed by spark.

Mr_Monkey, reosarevok : Can it possible that one work exist in many editions?

Yes

Any Shakespeare play for example has a million editions :)

It's even very likely

pristine_: nice!

iliekcomputers: would try to push it asap. A few changes are required.

Thanks :)

no worries. :)

pristine_: we should first discuss the format of the outputs of the queries that we will send over rmq

iliekcomputers: how would that be like?

pristine_: i have no idea yet :D

How about constructing a dictionary from the Dataframes and then sending over after converting to JSON?

hmm, yeah. but the dict that we send over will need to have a defined structure

Umm. We could have list of dicts where each dict would contain all the listen data of a user.

yvanzo: did you see my question about the nested variable tingies?

*thingies

(edit.display_data.${edit.display_data.entity_type})

pristine_: you should look at how we save data already using the stats module in listenbrainz

pristine_: I'd prefer to keep it similar

Nevermind, figured it out

iliekcomputers: okay :) I will go through it and would update you tomorrow.

yvanzo, bitmap: added https://github.com/metabrainz/musicbrainz-serve... - if you could help with the Flow question that'd be neat :) No hurry though, tomorrow's fine or whatevs

ok :)

alastairp: hey

[critiquebrainz] FannyVieira opened pull request #243: Update database schema (master...master) https://git.io/fhyuy

yvanzo: I wouldn't know how to add support for the id attribute off the top of my head, but it does currently output packaging, so it must know about it somehow?

Hello! While using "./develop.sh run --rm db psql -U acousticbrainz -h db " command to load psql session, i encountered the following problem "psql: could not translate host name "db" to address: Name or service not known".Can anyone guide me what to do next?

alastairp, iliekcomputers ^

reosarevok: still around?

bitmap: it does, but not packaging id

bitmap: my main doubt is that musicbrainz-server, mb-solr, and mb-rngpy (which sir depends on) rely on different versions of mmd-schema each.

yvanzo, hey

Maven Repository: org.apache.solr » solr-core » 7.6.0 (https://mvnrepository.com/artifact/org.apache.s...)

Basically you want to take a look at the compile dependencies here and make sure the common dependencies between this and mb-solr have same versions

You might have to do this a bit recursively.

For eg just looking over things, the version updates seem correct

Main culprit is slf4j generally

You want to keep that the same

Solr4j uses 1.7.24

I would also be careful about icu4j

Solr still uses 62.2

*62.1

I'd also avoid xmlunit update seeing as log4j is still using 2.5.1 https://mvnrepository.com/artifact/org.xmlunit/...

They way I do it is simply grep the jars in the solr download tar

I see, thank you.

About docker-solr, is it still used?

Now, we can probably avoid this if we use the maven shade plugin or not use the dependencies included jar

Which was my plan but I never got around to it

The latter is the recommended way

Release v3.0: Merge pull request #30 from metabrainz/test · metabrainz/mb-solr (https://github.com/metabrainz/mb-solr/releases/...)

Basically we should be using the normal jar

Not the one bundled with dependencies

Docker-solr, I still use it for Dev and testing

I think zas had some plans to move solr cloud setup to docker

It might also be useful for the vm

Both dockerfile and compose file are more up-to-date in mb-solr

As in?

I would like to push versionned images of sir and solr for use in the VM indeed.

Also mb solr build are failing

Well I last updated everything to 7.5

That should be good enough

https://github.com/metabrainz/docker-solr uses solr 6.6 whereas https://github.com/metabrainz/mb-solr uses solr 7.5

Ah!

Docker solr

Completely forgot about that

Yeah you can deprecate that repo

I already prepared a patch about it but realized it might be not worth it.

I merged everything in mb-solr

It was getting too fragmented

I think I haven't touched docker solr in over an year

Ok, I will just wait for zas confirmation before removing docker-solr, just in case he used it for development or something.

Ah no no

He's also using mb-solr

I thought you meant the dockerized solr in mb-solr

Not the repo

That repo is useless

Ok

samj1912: One last thing is about search indexes regarding latest changes to mmd-schema.

Yeah?

What exactly are the changes

I don't recall

mb-rngpy refers to a one-year old version, that doesn’t matter?

Have these been deployed already?

Okay, so there are a couple of places thing would need to be changed

Mb rngpy for sir to post things correctly to solr

Sir.wscompat probably to popular mbrngpy classes

*populate

And mb-solr

This thing

And probably solr schema if anything changed with the search Schema

Like new fields that were added which we want indexed

Or others that were removed

This changes will also go in sir.schema

Also obviously jaxb bindings will need to be updated

Ok, can reindex be made gradually or does it require to build and replace the whole index for related entity types?