Feedback and support thread for the Pulsewidth a-tisket instance

I’ve just been able to replicate it on a different release. Will investigate.

Edit: I thought I experienced the same issue but I can’t reproduce it. Strange.

What should you do with releases that barcodes that start with 00 like this one?

https://atisket.pulsewidth.org.uk/?cached=00602458615319-d_0-s_4SjzjaFsXvXiS7quZFzYEl-i_0

602458615319
or
00602458615319?

Someone already added this with the longer version and a-tisket couldn’t find it and warn that its a duplicate: Release “2 RUFF, Vol. 1” by Chase & Status - MusicBrainz

the itunes release linked there has a different barcode so it shouldnt be there but that’s another issue where it shows up as undefined in itunes if you add it manually, i checked with the script and it has a different barcode but it also has 00 at the start so now i’m stuck on what barcode it should have.

there was just recently a big edit note discussion about this

https://musicbrainz.org/edit/105112178

I don’t know where I stand on it personally, but I’d probably note which way you go in the annotation or edit notes

2 Likes

oh jeez… it might be valid then… :exploding_head:
either way, a-tisket cant find em.

Random question: The server caches the results of everything, right? Would you be able to use the cached results to generate statistics on availability of releases by country?

1 Like

Here’s an example of the data cached by a-tisket:

Aside from this Spotify API responses are cached because it requires authentication.

There is a bug in this tool:

It does not take all track artists: https://musicbrainz.org/edit/74168635

If I am in the wrong topic, please forward (I have muted digital-media, among others).

This is a known issue with the Deezer API. The best way to avoid it is to use Spotify or iTunes* as a primary source.

*As of this writing, a-tisket is not returning anything from the iTunes API, not even if a release ID is provided, so this option (is at least temporarily?) no longer available.

1 Like

The importer should blacklist deezer and allow only the sources that will make correct releases.
I don’t think the users will stop importing broken releases, otherwise.

FWIW, the current version of a-tisket picks up these additional artists. I don’t know whether marlonob’s instance did not do this back in 2020 or whether the editor was just too lazy to select all the correct artists?

4 Likes

Interesting! So not really practical to dig through it all to build statistics then? Not a big deal, I was just curious how often each country is represented. See the most included and most excluded countries.

Deezer’s API I think has just gotten better. They also now show copyright info on their page (don’t know about the API). When that release was added, those other artists weren’t mentioned by Deezer. Now they at least show on the webpages “featured artist” column.

It is? I’ve never had any issues with Deezer and multiple track artists.

I assume was a temporary issue with the API as nobody has reported any problems here and it worked OK when I tested just now?

So you found some errors in a 3 year old edit that was submitted from a completely different instance of the tool, you didn’t do any testing to confirm if this issue is present on this instance (which it isn’t), you don’t actually submit digital releases yourself and actively ignore discussion about them, yet you think it’s appropriate to jump into this thread to tell me to disable a third of the sites functionality? I thought your username was tongue-in-cheek but now I’m not so sure…

It is possible to do something like this but I’m not sure how useful and/or interesting the data would be. There’s likely to be a strong selection bias given that most of the people using the tool are looking at music available in the US / EU, so releases not available in those territories would be significantly underrepresented.

I can share the cache of Spotify JSON API responses with you if you want? It currently consists of just over 242k files (which should pretty much equate to the number of releases) totalling ~1.7GB. I wouldn’t be surprised if they took up less than 150MB when compressed though.

1 Like

For demonstration purposes I looked up this release in a-tisket with Spotify and iTunes disabled, and as you can see it omits any artists listed after the first ones.

It’s been flickering in and out all day, at least for me.

It appears that the API isn’t returning results for Apple Music on brand new releases. I just tried add https://atisket.pulsewidth.org.uk/?preferred_countries=gb&itu_id=1706916897&upc=5054197790454&preferred_vendor=itu&search_itu_countries=si

Released on 2023-11-10. This is the same thing that happened when the API quit working with barcodes. You can still get returns on the barcodes before that date. Hopefully it’s a short term glitch, but when I run any releases before about a week ago, I have no issues.

Update: It’s not just exclusively new releases either. Some are working, some aren’t. I just tried this and it’s returning no Apple results: https://atisket.pulsewidth.org.uk/?preferred_countries=US&itu_id=351168598&upc=016861779245&preferred_vendor=itu&search_itu_countries=si

2nd Update: When I run it on marlonob’s original, it works: https://etc.marlonob.info/atisket/?preferred_countries=us&itu_id=351168598&upc=016861779245&preferred_vendor=itu&search_itu_countries=si

3rd Update: Only works on the older Rob Zombie release. marlon’s crashes on the new release.

1 Like

Sorry mate.
I don’t understand atisket so I posted in the only unlocked topic I found.
I didn’t know there were several… Instances (?).

I didn’t notice the edit date but I also didn’t find a bug ticket list to see if this bug was known or fixed.

As you say I don’t know anything in this area.
It’s true but I preferred to report because I know this tool allows to add releases very quickly, do they better be good because some editors don’t really check the stuff they add.

I don’t really understand “tongue in cheek” exactly but yes my name was a joke. I wanted a name with a “has been” feeling and found this in 1 minute.

If it isn’t much trouble are you’re willing to put a copy up some place, maybe others might find that useful as well (This is a site of Metadata fans after all). I’m really bad at that stuff, but I might try to do it some time this winter for curiosity sake. But it isn’t hugely important!

a-tisket will need to be updated to remove Uruguay from the list of available Spotify countries after the pull-out takes effect, either on New Year’s Day or sometime in February 2024 (the article is not clear which).

1 Like

I thought this was sorted, but maybe not - Atisket still considers British Indian Ocean Territory as a country (which for all purposes it really isnt it just composes of a series of atolls and is sometimes inhabited by military personnel). This means that if Atisket doesn’t get a response for availability in this region it won’t set to [Worldwide].

At the moment I will manually override but it would be cool if it could do it automatically.

Example album: Spotify

Any chance of validating release dates / blanking out release dates that are clearly out of range? i.e.,

  1. Apple IDs are assigned sequentially. Assuming a label doesn’t put up a release for pre-order for years in advance, then it is possible to gate a range of +/-1 years around Apple IDs. For example,

    • 1720084561 is from 2024
    • 1683512766 is from 2023
    • 1633045198 is from 2022
    • 1585610528 is from 2021
    • 1509594482 is from 2020
    • 1477420592 is from 2019
    • 1358949269 is from 2018
    • 1221204617 is from 2017
    • 1180462915 is from 2016
    • 1001890192 is from 2015
    • , 811769944 is from 2014
    • , 681088393 is from 2013
    • , 509026157 is from 2012
    • , 484756845 is from 2011
    • , 380639200 is from 2010
    • , 327062805 is from 2009
    • , 160213834 is from 2006
    • , 102445265 is from 2005
    • the iTunes Music Store didn’t even exist before 2003
  2. When Spotify says a release date is January 1, it’s usually junk data.

Some users of the tool don’t think they need to think about this, so ideally the tool should be able to do it for them. (Interestingly, the YouTube videos backing YouTube Music have upload dates, but I think only the month precision can be relied on since the displayed date can differ from more established sources, even after accounting for correcting the timezone.)

By the way, something that can’t be accounted for is different markets having different release dates. (For example, many Japanese releases were only available in Japan, and then international markets were added to an existing release years later. In those cases, Japan should have a different release date than whatever people are trying to use for worldwide these days.)

8 Likes