Now that we have more explore tiles in there I wonder if we should simplify the images anyway, it’s looking a bit messy (but also fun and colourful, I guess), if you agree I’ll make a ticket
2024-01-26 02658, 2024
aerozol
outsidecontext: I posted about 2.11 on the socials, hope that’s okay 👍
that seems like trying to reverse-NGS the rg groups
2024-01-26 02614, 2024
ApeKattQuest
huh. that's a really boring and grey cover for an album called "in rainbows"
2024-01-26 02607, 2024
ApeKattQuest
reminds me of the idea that all tracklists of every release where merged *unless* they had adiffernt tracklist. so you had one release that was actually 6-8 releases (with only release events attached + disc ids) and thne maybe one or two This thing has extra tracks" typ releases as well
2024-01-26 02622, 2024
ApeKattQuest
it was amess and i'm SO happy we don't do that now
2024-01-26 02601, 2024
ApeKattQuest
but there eare *still* release out there ewith merged disc ids that really shouldn't be and I wish there was an esy linkable page doc that explained this exact thing
2024-01-26 02621, 2024
ApeKattQuest
is due to historically radical different way of storing the data
2024-01-26 02636, 2024
ApeKattQuest
so it's OK to move/remove discids
2024-01-26 02644, 2024
ApeKattQuest
is there evn such a thing? reosarevok ?
2024-01-26 02640, 2024
reosarevok
A doc about crap discids? Honestly, not sure
2024-01-26 02637, 2024
ApeKattQuest
I mean
2024-01-26 02642, 2024
ApeKattQuest
they werent crap originally
2024-01-26 02650, 2024
ApeKattQuest
they weren't added to one tracklsit
2024-01-26 02600, 2024
ApeKattQuest
what hpapened was that all tracklsits were merged/added to one thing
2024-01-26 02638, 2024
ApeKattQuest
and thne whne ngs came this was solved by splitting those int odiffernt releases but then assigning every single one of thme every discid
2024-01-26 02634, 2024
ApeKattQuest
because there wasn't (and still isn't, rereleases can share discid) any discid<->release-event link
2024-01-26 02630, 2024
ApeKattQuest
when we coded that originally. man we were naive about rl musicreleases database mapping :DDD
mayhem: Moving here because its a better fit... What I didn't brought up yet is my suggestions to employ renovate in all the projects I touched to update the dependencies
2024-01-26 02646, 2024
mayhem
k.
2024-01-26 02624, 2024
mayhem
should I merge the PR and then fix tests?
2024-01-26 02625, 2024
mayhem
merged. starting on tests. well, after lunch.
2024-01-26 02654, 2024
gcrkhehim[m]
You can review and merge the PR; however I don't think the tests need to be done NOW, its just something to consider
IMO raw images should be ignored unless they are the last possible choice to obtain a front cover for an album.
2024-01-26 02624, 2024
kellnerd
i.e. for the whole release group
2024-01-26 02613, 2024
mayhem[m]
gcrk (he/him): the content resolver is all I have been working on. at least the results are good. :)
2024-01-26 02617, 2024
gcrkhehim[m]
I mean there is nothing left stopping me from doing the packaging work
2024-01-26 02629, 2024
mayhem nods
2024-01-26 02630, 2024
gcrkhehim[m]
does it even make sense to bring it to pypi?
2024-01-26 02643, 2024
gcrkhehim[m]
If its a enduser tool, the pyinstaller is probably the important bit
2024-01-26 02637, 2024
mayhem
yes, it does. not wholly-end user -- for instance I've started yet another project, called LB-local. that is to showcase all the features of the resolver/troi and get playlists uploaded via subsonic.
2024-01-26 02625, 2024
mayhem
we're going to explore the possibility of making that available as web components for easy inclusion into other apps. but that is daydreaming right now.
2024-01-26 02653, 2024
mayhem[m]
gcrk (he/him): I invited you to the matching tools repo.
2024-01-26 02619, 2024
monkey
kellnerd: Currently, the image used is the one that is set as the selected cover for the Release Group on MB. If that's not the desired outcome, i think the solution should probably be to set the digital media cover on MB.
2024-01-26 02619, 2024
monkey
IMO we probably shouldn't override the community's choice of most representative cover with an automatic process.
monkey: Weird, the RG cover was automatically selected and has not been explicitly set. So I guess this would rather be an MBS issue instead :)
2024-01-26 02625, 2024
monkey
Hm
2024-01-26 02612, 2024
kellnerd
What I still don't understand is why the LB album page does not use the same cover as the artist page cover grid.
2024-01-26 02602, 2024
monkey
That I can't tell you. I think because we use an automagically selected "canonical" release and their artwork. In this case the selected canonical release is the digital album
2024-01-26 02624, 2024
monkey
For the album pages, we use the RG's selected cover art directly I think
2024-01-26 02620, 2024
monkey
Looks like the first edit on MB adds the raw cassette photo, probably why it was auto-selected as the RG cover art: https://musicbrainz.org/edit/88609953
it's ordered by release_group_cover_art.release (should have been NULL in this case, since nothing was explicitly set) and then by release date (same for both releases), so I guess it is arbitrary after that
2024-01-26 02635, 2024
bitmap
since it only checks is_front and not any of the other types
2024-01-26 02641, 2024
monkey
Right, I think the key in this context is that `is_front = true` is true of both of the images in question. Ons is "front" only while the other is …
2024-01-26 02649, 2024
monkey
yep, I'm too slow, you said it :)
2024-01-26 02602, 2024
Lotheric
🦄
2024-01-26 02657, 2024
twodoorcoupe has quit
2024-01-26 02627, 2024
test1234 joined the channel
2024-01-26 02625, 2024
test1234 has quit
2024-01-26 02607, 2024
kellnerd
Ah, so it is random and MB editors should pay attention whether the RG cover art changes in an unexpected way when they add a new front cover.
2024-01-26 02630, 2024
kellnerd
Or is there a simple fix to ignore raw images for RG cover art?
2024-01-26 02617, 2024
bitmap
it should be fixable with another order by clause (e.g., a case statement checking if index_listing.types contains raw)
2024-01-26 02607, 2024
gcrkhehim[m]
<mayhem[m]> "gcrk (he/him): https://github...." <- I don't think the testing touches the deployment, but the ci pipeline needs to be adjusted
2024-01-26 02618, 2024
gcrkhehim[m]
I can port the reporting from the other repositories now
2024-01-26 02659, 2024
mayhem[m]
thx!
2024-01-26 02622, 2024
mayhem[m]
🦄 contract signed, waiting for counter-signature before going public.
2024-01-26 02651, 2024
monkey
Woohoo ! 🎉
2024-01-26 02656, 2024
mayhem[m]
right-o. time to enslave some robots for purposes of making cocktails this evening. ta-ta!
2024-01-26 02627, 2024
monkey
ansh: made some progress with the save to image !It's still not perfect on mobile, not sure how to deal with the size discrepancies, but on desktop looking nice ! https://usercontent.irccloud-cdn.com/file/UdfnfyP…
2024-01-26 02637, 2024
ApeKattQuest
I'm getting A LOT of "search error" whne i try searching recently
2024-01-26 02655, 2024
ApeKattQuest
in general every search does this and a refresh thne makes it work
2024-01-26 02622, 2024
ApeKattQuest
btw on the JIRA is it possible to make the comments and the like NOT search "newest first"
2024-01-26 02623, 2024
ApeKattQuest
it had always sorted oldest in the past s oit makes it really confusing to read comments now (and also while you cna chose it it seems to revert back to "newesst first" every time i close pages)
2024-01-26 02638, 2024
ApeKattQuest
s/search/sorted/
2024-01-26 02600, 2024
minimal has quit
2024-01-26 02610, 2024
bitmap
ApeKattQuest: that annoyed me too, though it seems to have remembered my setting after I changed it once
2024-01-26 02639, 2024
relaxoMob joined the channel
2024-01-26 02641, 2024
SigHunter has quit
2024-01-26 02629, 2024
SigHunter joined the channel
2024-01-26 02641, 2024
aabbi15 joined the channel
2024-01-26 02636, 2024
wiso joined the channel
2024-01-26 02646, 2024
wiso has quit
2024-01-26 02600, 2024
ApeKattQuest
hmm maybe becuase I use both mobile and a differnt browser? idk
MBS-8393: Extend dynamic attributes to all entities
2024-01-26 02630, 2024
aerozol
Mōrena!
2024-01-26 02636, 2024
ApeKattQuest
hey aerozol !
2024-01-26 02633, 2024
aerozol
monkey: kellnerd: I think it should use the release group image everywhere in LB (or not at all). When that's not a good one it's easy to fix, at least. And consistent. If I was following the gist of the convo
2024-01-26 02644, 2024
aerozol
Morning ApeKattQuest!
2024-01-26 02654, 2024
monkey nods
2024-01-26 02625, 2024
aerozol
Did you see the LB Local logo monkey? What do you think, does it work with your MeB set?
