Talk:Q202698

Reversion of P155/P156

Hi @Moebeus:, how are you ?. I understand this reversion; the song is not the track. Sorry, I follow the sitelink of the WP article that talk more about track than the song. Is it correct to use the published in (P1433) of the track item to show the list of albums ?. See the record of this song. Your suggestion to get info from tracklist of album is not possible from WP, because backlink recovery is not available (as far as I know). Thanks, Amadalvarez (talk) 05:35, 18 April 2022 (UTC)

Hi there!

Using "published in" on the track is really redundant, but less wrong, if that makes sense. It's sometimes used as a visual aid, but it really shouldn't be there at all.

I strongly recommend installing the following user script if you like to edit music:

mw.loader.load("//www.wikidata.org/w/index.php?title=User:Nikki/ExMusica.js&action=raw&ctype=text/javascript");

This will make it a lot easier to visualize how the music model works.

As for how to get at this data from Wikipedia I really don't know, but if you have a specific use case I can ask around if you want? Lots of people much smarter than me that might know. Or come join our music editor chat on Telegram: https://t.me/exmusica

Moebeus (talk) 11:01, 18 April 2022 (UTC)

Hi, @Moebeus.

Well, even though I'm working on retrieving WD information for the infobox, I won't finally put the recording and album information of a music composition inside the infobox, but in a table within the text of the article, built with the Listeriabot. The main reason is that, in popular songs, the resulting list is too long to be included in an infobox.

See the solution for song Yesterday and the case of track recorded by The Beatles. I do both cases, because WP articles sometimes talk about song and its versions, and sometimes from a performance and the albums where it is included.

Besides that, we have a few wrong structure:

• items with P31 = musical work/composition (Q105543609) + single (Q134556). The perfect solution should be split it in 2 items: composition and track. But, looking at the contents, the Q105543609 could be removed (by the way, in many cases, added by you ...)
• items with P31 = a kind of record + single (Q134556). In this case, I think we should be flexible, because if a composition has only been recorded to appear in a single and nothing else, it seems too strict to have 3 items to represent it.

What's your opinion about these two kind of errors?. Does it worth to clean them?. That doesn't affect to my work with infobox, but we may use this opportunity to fix them.

Thanks, Amadalvarez (talk) 19:13, 3 May 2022 (UTC)

• Q105543609 + Q134556 is *always* wrong. Luckily there aren't very many of them, but those need to be fixed - compositions and singles are separate things. And track is a third thing, also separate.
• As for your second point I'm not sure I understand what you mean when you say "a kind of record"  ?

Moebeus (talk) 20:28, 3 May 2022 (UTC)

Ahhh, sorry. "kind of record" is the different Qids I found that I understand as "record track" in the linked query:

SELECT ?item ?itemLabel  ?basat ?basatLabel WHERE {
:  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }
:    VALUES ?glam {
:                      wd:Q6942562
:                      wd:Q55850593
:                     wd:Q55850643
:                      wd:Q193977
:                      wd:Q106588852
:                      wd:Q7302866
:                      wd:Q63481999
:                      wd:Q6942562
:                      wd:Q106805967
:                      wd:Q107080092
:                      wd:Q47000326
:                      wd:Q64100970
:                      wd:Q20871935
:                      wd:Q11931373
:                      wd:Q25203386
:                      wd:Q677466
:                      wd:Q20077126
:                      wd:Q3302947
:                     }
:  ?item wdt:P31 ?glam;
:  wdt:P31 wd:Q134556.
: OPTIONAL { ?item   wdt:P2550 ?basat. }
:}
:

Try it!

Thanks, Amadalvarez (talk) 20:58, 3 May 2022 (UTC)

That query gives me an error "Query is malformed: Encountered " "service" "SERVICE "" at line 2, column 4.", I'm not very good at sparql or I would try to fix it myself :) Moebeus (talk) 23:10, 3 May 2022 (UTC)

The first one in the list - musical performance - Wikidata - is an academic profession, and as such has nothing to do with singles at all. If that shows up it's a clear mistake 🤔 Moebeus (talk) 23:14, 3 May 2022 (UTC)

@Moebeus: Forget the query, the problem are the colon (:) inserted by the wikieditor. But looking the definition about the P31 in Wikidata:WikiProject_Music#Track_properties, I re-focused the problem with 2 different queries:

• Items with P2550 with a wrong P31 (according to Wikiproject) >> https://w.wiki/584g
• Items with correct P31 + P31 = Q134556 >> https://w.wiki/584e

To sum up, our challenge are:

1. As you said yesterday, Q105543609 + Q134556 is *always* wrong. I propose try to split into 2 items if we have enough information to do it, or remove Q105543609 when there is not info about composer, creation date, etc....
2. Items with a wrong P31: I propose to review in which cases we just ​​need to add one of the correct values as a subclass of the P31 (eg: Q106805967?) or as a additional P31 of the item. Or else, simply change the P31 when wrong.
3. Items with correct P31 + Q134556. Note that there is only one case pointing to a composition (P2550), and there are very few with composer data, i.e. we are not in a position to create a composition item either. These are the cases where the performer is the author and the single is the recording itself. This is a similar situation to self-published video clips on Youtube that will probably never appear on any album. In fact, digital records have changed the concept of "distribution media" which is losing its meaning. These cases are not a problem for my work with infoboxes. However, in my opinion, should be allowed as long as it is not used in another version (album, compilation edition, etc.). In any case, without forgetting that the composition should always be an independent item.

I hope I have clarified the situation so that we can proceed. Thanks, Amadalvarez (talk) 15:48, 4 May 2022 (UTC)

@Moebeus I start with #1. We're below 100 !. Amadalvarez (talk) 05:18, 6 May 2022 (UTC)