#metabrainz

[bookbrainz-site] snyk-bot opened pull request #539 (master…snyk-fix-4a264384302d4afcca7d715616156c40): [Snyk] Security upgrade snyk from 1.385.0 to 1.425.4 https://github.com/bookbrainz/bookbrainz-site/p...

Project bookbrainz-site build #3564: passed in 4 min 41 sec: https://travis-ci.org/bookbrainz/bookbrainz-sit...

[musicbrainz-server] mwiencek opened pull request #1809 (master…editor-data-handling): Improvements to editor JSON handling https://github.com/metabrainz/musicbrainz-serve...

Mr_Monkey: ping

pong

this is how it looks now

Yeah, the flashing color is an issue

are you suggesting to change bg to white when text is follow?

on hover

That green orange color will go with the other PR

that is open

Well, if the button is blue background in normal state (not following), then it should change to something else when you hover it, yes

ohk thanls

thanks?

!

And since the way we're doing that with the other style, I guess it's the logical way to go.

makes sense. I was confused to keep blue or white bg for "Follow" on hover

yvanzo, bitmap: around? :)

hey

I am working on updating https://github.com/metabrainz/musicbrainz-serve... and hope to respond to some points made in the community forums asap, so probably not going to review stuff until that's done

but I can probably rubber-stamp some simple ones if you link them

Anything you'd want me to go through related to the forums and whatnot?

And/or this code

I'm mostly just responding to the question about the timline

and for the PR I'm seeing about reducing a couple uses of UnsanitizedEditorT

reviewing the checkForUnsanitizedEditorData function would be helpful, if you see a way to improve that further

So basically: if it's a dev build, every hydration should check every key for anything that looks like an editor, and make sure it only has the sanitized content, and just in case it also checks anything that doesn't claim to be an editor and looks for birth(date), email, password?

I don't really see anything else in

you've got it

That we really should look out for, since all the other data is public if it exists AFAICT

Well I guess preferences aren't, technically

yeah, it might make sense to check for preferences here

hi

bitmap: is 1801 ready for merge now or of lesser priority?

it is ready, I just haven't had time to deploy it yet (requires running scripts on floyd)

reosarevok: do/did you/we have any direct contact with BBC Music? would be interesting to know why they removed artist pages.

We kinda-sorta do, but IIRC they've been moving away from those for a while. I guess people weren't using them much? tbh they didn't have much content

I don't think it's anything with us or the data (as ruaok mentioned they still use MB in many places), just probably cost-cutting really

IIRC, the rel was already limited to pages with extra content, so they did show extra content (interviews, lives, samples)?

Well, they sometimes did, but by far most of them did not :)

Apparently a few redirect to kinda "tag" pages

So if you go to what used to be the page for Adele, now it just shows you all the articles for Adele

Most don't do that either

bitmap: I guess we need 1801 to run for a couple of days on beta before releasing anything then?

I guess it might not matter - the triggers will be in the DB either way, right?

So either it makes a mess in both again or it works :D

let's try not to release something that makes a mess again :)

Well, it's been tested and seems fine, but the other had been tested too and it struggled with real editing

But I don't think we have a better way of doing this

We just need to release when we're around to revert if 504s come back :)

(and I guess have the reverting script ready just in case)

I agree on "we [don't] have a better way of [testing] [the trigger] [with real editing]" but it still does not require to update the production branch with and to make a server release.

Oh sure, my point is that releasing the trigger will affect prod anyway

Not suggesting releasing the whole thing to prod necessarily

Yes, when I wrote "releasing", that was about MBS production branch.

But since that's the only potentially breaky thing, it's probably not much of a difference if we just release after a few hours of not breaking

Since if it does break we should notice fast :)

It makes a difference if "it struggled with real editing" again.

The only thing that did that was the one released on all servers though? But anyway, sure, let's test it for a couple days