Not necessarily. If they can find matching images online (Amazon or elsewhere), they can upload those just fine without being anywhere near a scanner. Yes, that user says uploading cover art is bothersome, but it improves the project much more than using Amazon CA does, since we can actually do things with images in CAA.
I didn’t quote/say that, au contraire. Please re-read it.
Edit: I’ve run the numbers, and around 18% of our releases have ASIN and no CAA image (~17,88% has no CAA image, ~18,06% has no CAA front image). How many of those actually resolve though?
SQL used for anyone wanting to double check the numbers
SELECT COUNT(DISTINCT musicbrainz.l_release_url.entity0) AS releases_with_asin_and_no_caa
,(SELECT COUNT(*) FROM musicbrainz.release) AS all_releases
,(COUNT(DISTINCT musicbrainz.l_release_url.entity0)*1.0/(SELECT COUNT(*) FROM musicbrainz.release)*1.0) AS percentage
LEFT JOIN (
SELECT DISTINCT release FROM musicbrainz.cover_art /*WHERE id IN (
SELECT id FROM musicbrainz.cover_art_type WHERE type_id IN (
SELECT id FROM musicbrainz.art_type WHERE name = 'Front'
) caa_releases ON musicbrainz.l_release_url.entity0 = caa_releases.release
WHERE caa_releases.release IS NULL AND link IN (
SELECT id FROM musicbrainz.link WHERE musicbrainz.link.link_type IN (
SELECT id FROM musicbrainz.link_type WHERE link_phrase = 'ASIN'
Once we have CAA Artwork for all releases that have ASIN Artwork then that would seem a good time to remove the Amazon artwork and not before, I cannot understand why we would contemplate before then.
The fact is that it is not easy to add data to Musicbrainz in bulk I have lots of high quality data that I would like to add to MusicBrainz to improve it but it is not possible to add it in an efficient matter, it would take me many years of fulltime work to add to MusicBrainz using the prescribed method therefore I am unable to add it.
And Im sure this would apply to adding artwork for all those Amazon releases as well.
Removing a useful feature to ‘push’ users to contribute is not putting the user first and that can hurt the service in the end.
If the aim is to get more people to upload covers then maybe allowing people to tag with them but putting a watermark or a layer over the image on the release page would help.
I would be supportive of adding more ‘auto’ image grabs from other sources if it makes the service better for the end user.
If there’s another good reason for removing the images then have at it
Who’s the end user? The people browsing the website? I can see that argument. But what about users using Picard (or beets or other software) to tag their files? The displayed Amazon cover art doesn’t mean that Picard can fetch Amazon cover art (or that it will even be the same) and relies on Picard having the Amazon CA module enabled to begin with. People using the web service? They can get a list of links and from that get an Amazon ASIN and themselves try and deduct the Amazon CA (if any), but even the most basic web service response tells whether a release has CAA images. Same for data users using the database directly - figuring out Amazon (or other sources’) CA is a lot of extra steps, but the database directly contains information about where there’s CAA images (as well as details about them).
Amazon CA isn’t being using on CritiqueBrainz, and I don’t think anything but CAA is going to be used for ListenBrainz. (AcousticBrainz currently doesn’t utilise cover art, but I also don’t think they’d use anything but CAA.) This means that there would be an inconsistent user experience when hopping between the various *Brainz sites—which is one of the things we/MetaBrainz want to encourage people to do more.
Anyway, regardless of why I would like to see it dropped, it seems like most people here would like to see it remain for the time being, so I’ll be sure to bring that feedback to the meeting tomorrow.
Thanks for listening. I certainly agree it is wrong to expand a depreciated artwork system to other parts of MB. But maintaining the limited art where it is sounds good.
Please consider adding text to the CoverArt page pushing people to upload replacements for Amazon artwork. I have seen some releases with insanely high resolution scans of the rear and booklet, but no fronts.
People need nudging to replace the Amazon art as most people don’t read these forums they won’t know that it is in danger of being removed.
You have nudged me to handle art different now… if I see an automatically linked bit of art coming from Amazon now I’ll just go grab that art and dump it on CAA instead. Personally I don’t really like doing this as I don’t have the “CD in hand” to always confirm, but it is in the aim of not loosing what we have now.
Perhaps this could form the basis of one of the “community cleanup” initiatives? Check and upload to CAA if appropriate. I do believe that the goal should be to do away with the linked cover art from Amazon.
In echoing what everyone is saying, I am very for just letting this trend naturally develop so that the amazon % keeps decreasing until it becomes negligible and eventually the “turning it of” will seem logical to everyone since it’s (practically) not in use.
That being said, I also don’t see anything wrong with “depreciating” it and f.ex not fixing any trivial bugs (huge-database-breaking ones are a different story obviously) by closing bug-tickets with “amazon is generally depreciated, we encourage uploading to the CAA as a fix”
I also agree with jesus2099’ and IvanDobsky’s suggestions: make it obvious these are amazon and maybe incorrect, encourage people to upload to the caa instead. (this is also relevant to our re-designer, chaavi!)
Other than for the first few minutes (where the IA is processing (e.g., making thumbnail versions) the uploaded images), this shouldn’t be the case. Consider opening a ticket or a new forum topic so what’s going on can be debugged/sorted out.
Could we at least make it more obvious that a cover art is provided from amazon and therefore can’t be trusted? Like a red border around the cover, a warning symbol besides the “Cover art from Amazon” text + adding “(may not be accurate for this release)” or hiding the cover behind a translucent white haze which disappears if you click on it?
5% of artwork is a big hit, so it is a pity there is no announcement being made beyond a note in some meeting minutes. My views have changed since the start of this thread as i know how dodgy that art quality can be. though this thread had triggered me back then to go on a chase of artwork for many of my subscribed artists that were heavily reliant on Amazon art.
it would be good to get an announcement on the front page as a nudge to get a bit of a push to grab some more art before the images are turned off. or at least change the tiny “cover art from amazon” title to a louder bigger warning prompt
Edit: Lights Go Out = Images Turned Off. Guess it didn’t translate well. Sorry for more confusions.
Right now there is <5% artwork from Amazon being displayed due to a change in Amazon’s API. The decision was not so much made to remove the support because we no longer want it, but rather to remove the code instead of trying to jump through hoops to change it to adhere to Amazon’s new API.
The decision was made 3 days ago, at the weekly meeting (which everyone is welcome to join and chime in at) – of which I try to have the notes posted here on the forums within 24 hours. The decision having been made does not mean it gets implemented right away (in fact, this seems like a fairly low priority thing). The posting of the meeting notes is not an announcement in and of itself.
Again, the change hasn’t been made/implemented yet. At the very least it will also be announced in the MusicBrainz server update notes once the change has actually been made. I don’t know if it really warrants its own separate blog post, but maybe we will do that too. We’ll see. Besides, as you may know, changes to MBS code does very rarely happen overnight. There’s a good chance the broken Amazon code will remain for a while as the MBS devs focus on working on other areas of the code. MBS-11848 may also get made and sit for some time before the faulty Amazon cover art code is finally dropped.
We won’t be removing Amazon links from the sidebar, let alone entirely. You can still go to the Amazon page from the release page and get artwork. Nothing’s changing there for the foreseeable future.
Just for fun/reference, the stats of CAA cover art vs. Amazon cover art coverage for the last ~2 years.
I was not trying to be negative. The 5% figure came from your meeting notes which are always interesting to see. And it is clear in there that this is an Amazon change and no point in wasting MB time on keeping up with them as they will keep changing. I was trying to suggest it can be spun to a positive and use it as a nudge to people to grab artwork for their favourite artists.
I just want to additionally note that this change has no impact on the working or not working of Picard’s Amazon cover art plugin. That’s purely depending on a release having a link to Amazon, and on Amazon’s URL structure not changing.
MBS-12200: Drop schema objects related to Amazon cover art support . For a long while, releases with Amazon URLs would be checked for cover art on Amazon, and if found, a link to the image would be cached for display. Unfortunately Amazon’s API to do this changed, and we haven’t synced artwork from them in years. We still have many old images cached and we still display those, but they aren’t guaranteed to be in sync. Last year we decided to drop support for displaying these, while giving time for users to upload any correct images to the Cover Art Archive. To help with this, we have a report of releases with Amazon cover art but no Cover Art Archive front cover.