My concerns regarding this was that, I cannot automatically distinguish between data provided by the database item behindPlaydate Catalog ID(P12125) and information that just happens to be found under the same url.
Let's say the footer of a page might contain an email address, that is not related to the person reviewed in the page.
Now if I had an indication that a url that resolves to (for instance) anIMDb ID(P345) usually holds the information that applies totitle(P1476), then I could make the assumption that it would make for a fitting reference statement when extracting this property from the url.
This information could be useful for other tools like:
A tool that tells you that an item should have a title and that you could find it on the IMDB
A tool that highlights a reference that might be faulty because it occours in an unusual context. Let's sayDeku Deals ID(P8364) is used on adate of birth(P569) statement.
If an ID points to a wiki, that could hold all kinds of statements, it should be set tounknown.
Support I quite like the idea, and it falls in line with what I am trying to do with more visual cues on items for the viability of references / sources and notability·addshore·talk to me!10:23, 28 October 2024 (UTC)[reply]
Wait this property mixes three things together (1) Some entries in the database have data on this property. (2) All entries in the database have data on this property. (3) All entries have reliable data that's worthy of being imported on this property.
I read the proposal and all comments again and I'm left with the impression the only assumed meaning of this property is (1). (2) can be expressed with a qualifiernature of statement(P5102) →sometimes(Q110143752). (3) if there is a property for it, I'd assume that the is a consensus that it is worthy of importing 🤷
A tool that tells you that an item should have a title and that you could find it on the IMDB does not mean you shouldn't review it first. Did I imply we should do automatic imports based on this property alone? –Shisma (talk)11:48, 2 November 2024 (UTC)[reply]
@Shisma the general meaning of "should" is "there may exist valid reasons in particular circumstances to ignore a particular item, but the full implications must be understood and carefully weighed before choosing a different course".
While you didn't explicitely suggested automatic imports, the wording that you did use encourages people to read it as a justification of automatic imports. If you propose a property you not just have to think about how you want to use the property about also about how other people might misuse it and think about how to design the property to prevent misuse.ChristianKl ❪✉❫09:31, 7 November 2024 (UTC)[reply]
If the data provided is faulty, wrongly formatted, or unreliable we can surely express it with ranks and qualifiers. Don’t you think? —Shisma (talk)08:49, 12 November 2024 (UTC)[reply]