Template talk:Lingua Libre record

Replace by SVG

{{Edit request}}

Please change Crystal Clear app download manager.png by Crystal Clear app download manager.svg. Thomas Linard (talk) 07:26, 30 August 2021 (UTC)

 Done in Special:Diff/589602199. --TKsdik8900 (talk) 08:09, 8 September 2021 (UTC)

Move content of the template to Structured Data on Commons?

Shouldn't we move the whole content of this template and the file description pages (maybe apart from parts of categorization) to Structured Data on Commons? It seems more than much else predestined for it to me. I see that SDC is already used, but is there anyone familiar with the process of file creation by Lingua Libre and could say if it would be feasible to put the data directly in SDC without putting highly structured file specific content in wikitext? Best, --Marsupium (talk) 04:51 & 04:54, 22 May 2022 (UTC)

@Marsupium: There is a Phabricator ticket that tracks progress on this task. Now it is stalled, mainly because there is no Wikidata equivalent for the Lingua Libre properties language level (P16) and place of learning (P15). That's said, Lucas Werkmeister has made some work recently on that topic. Pamputt (talk) 14:00, 7 April 2023 (UTC)

@Marsupium and Pamputt: Oh, I wasn’t aware of this thread or that task. Yes, Nikki and I were working on this – the template/module can now pull all its data from structured data, and NikkiBot is currently migrating tons of files to this format (random example). Lucas Werkmeister (talk) 16:06, 7 April 2023 (UTC)

About Language level, we can reopen the former request for a wikidata property, but based on well known and international en:Common European Framework of Reference for Languages levels. Yug (talk) 17:34, 7 April 2023 (UTC)

Sounds like a reasonable way forward for me. Pamputt, Lucas, Nikki, what do you think? —Marsupium (talk) 15:54, 8 April 2023 (UTC)

Ok for me. Pamputt (talk) 07:03, 10 April 2023 (UTC)

I think a language level property could be useful for other things (e.g. as a qualifier of languages spoken, written or signed (P1412)) but I wouldn't add it to Lingua Libre files here. It makes more sense to store it once in Lingua Libre rather than duplicating it on every file by that person. The structured data already contains the Lingua Libre ID if people want to fetch that information (here's an example query for the file Lucas linked above: https://w.wiki/6ZvZ). - Nikki (talk) 12:48, 11 April 2023 (UTC)

@Nikki There is an interest in "duplicating" the information: this information can change over time. For example, a speaker may say that he is a beginner in a given language, record some words a few years later, change their language level to intermediate and record new words. If we only rely on the current status of the Lingua Libre Wikibase, the language level information for the first recordings will be lost, unless we search the user's history. Now I am thinking that Lingua Libre should use the start/end time as a qualifier to track when the user changes his statement. Any thoughts? Pamputt (talk) 09:10, 15 April 2023 (UTC)

Agree. --Marsupium (talk) 11:01, 28 April 2023 (UTC)

That still doesn't sound like a good reason to me. If someone's proficiency level changes, that can be modelled using start/end dates for each proficiency level without needing to copy it to every single file they ever record. - Nikki (talk) 07:10, 8 January 2024 (UTC)