Wikidata:Requests for comment/Creating properties which link to an archived website

An editor has requested the community to provide input on "Creating properties which link to an archived website" via the Requests for comment (RFC) process. This is the discussion page regarding the issue.

If you have an opinion regarding this issue, feel free to comment below. Thank you!

It is very unclear whether we should create new properties that have formatter URLs which link to the Wayback Machine (Q648266) or other archival sources.\

In addition, there is precedent but no official policy whether we should change formatter URLs of dead properties to be to the Wayback Machine (or to simply delete the properties).

For example, see this proposal for a sports-related identifier or this PFD request for a property whose formatter URL went offline.

(This is my first RFC, please be nice ) —Tomodachi94 (talk) 18:01, 22 April 2023 (UTC)[reply]

Discussion edit

If the former database has well-archived links, then why not use them? Here is a good example of such database. Regards Kirilloparma (talk) 17:58, 26 April 2023 (UTC)[reply]
  • From my point of view, the usefulness of archived identifiers is limited by the fact that they can become outdated, but this will depend on the nature of the external database and the way it's used. If the archived database still provides useful information, it's probably OK to create a property linking to it. BEANS X2 (talk) 13:01, 2 May 2023 (UTC)[reply]
  • I see two problems. First, an archived url property would need to either link to the general archive.org page for that url or link directly to an archived version. The later would tell you when the person posted the link, but the general calendar link doesn't tell you much at all and you have to hope the site has been archived. Second, when should an archive link be used and when should a regular url be used? Both would be useful in different ways. I think the ideal, pie-in-the-sky solution would be to have an integration with the wayback machine directly that would ping the machine to archive the url when it is added to wikidata. The url property would have some kind of automatic sub property that is a link to that archive, and the date would be added with a property specifying it is the date of the archive. That may be too much to do, the other solution could be having an archive.org integration that simply tells the reader if the url has been archived and the latest date if it has been. Ha2772a (talk) 18:38, 18 May 2023 (UTC)[reply]
  • I don't see any arguments for why we should create new prorperty that link to archived websites. As far as existing properties go, data stability is an important value of Wikidata and thus we strive to solve issues in a way where we don't delete data. While we don't have a policy about whether to "simply delete the properties" we do have a process for deciding whether to delete properties and that process is usually against deleting these kinds of properties if there's significant data in them. ChristianKl11:43, 17 June 2023 (UTC)[reply]
  • Wikidata brings a lot of value by being a Rosetta Stone in the Linked Data Web through its huge number of external identifiers. Storing some widely-used but discontinued identifiers can have value in at least the following scenario: I have a dataset that contains the discontinued IDs and I can use Wikidata to map it to another identifier that is not discontinued, thereby easing my migration. While I don't think we should necessarily create new Properties for deprecated websites I would advocate for not deleting what we have. --LydiaPintscher (talk) 19:02, 17 June 2023 (UTC)[reply]