Jump to content

Community Wishlist Survey 2020/Wiktionary/Adopt Lingua Libre Bot service as a tool

From Meta, a Wikimedia project coordination wiki

Adopt Lingua Libre Bot service as a tool

  • Problem: Lingua Libre is a great service to record pronunciation of words and, now, Lexemes at Wikidata. When you record them they're uploaded to Commons and via Lingua Libre Bot they are added to the corresponding word/lexeme. But this bot is mantained by a volunteer, and it seems that sometimes it can be stopped for weeks. This service should be adopted as a WMF Tool and make it more stable and not dependant from an user disponibility.
  • Who would benefit: People wanting to record pronunciations
  • Proposed solution: Add it to Toolserver and run it independently.
  • More comments:
  • Phabricator tickets:
  • Proposer: Theklan (talk) 11:01, 27 October 2019 (UTC)[reply]


IMHO, the problem is more general. On a long-term period, LinguaLibreBot and Lingua Libre website itself has to be maintained to add more feature and to fix bugs. I think Wimedia France starts to think about this problem (Eavqwiki may comment more). Pamputt (talk) 17:20, 1 November 2019 (UTC)[reply]

Allready an grant, although the solution is different. Will use OAuth, so no bot will be needed. On the talk page, the developer writes: "the files will be transferred from LinguaLibre to Commons using an OAuth authorization". So, in the future an bot will not be used and this issue will be fixed in another way than the proposee here accounted for. (Grants have a lifetime of 12 months)--Snaevar (talk) 12:37, 11 November 2019 (UTC)[reply]

@User:Snaevar The project supported by the grant has been implemented: the recordings are automatically uploaded through OAuth on Commons as Theklan mentioned, when it used to be up to the developer to transfer them periodically from the beta version of lingualibre that didn't use OAuth. Even with OAuth though, the bots are needed to upload the recordings to specific places on other projects - not just Commons -, that's where Theklan suggests it'd be useful to have the tool internal to Wikimedia to internalise both the recording process and the "storing" process. Correct? I think that having a separate Lingua Libre website and recorder is useful because that's where contributors without a Wikimedia account can also record (using a wikimedian's account for now but hopefully not anymore in the future). It's easier to explain the use and value of audiovisual content for languages' vitality to a greater number of speakers on Lingua Libre, all the while uploading the ultimate content on Commons - as WikiTongues does for example. Wikimedians on lingualibre could then validate non Wikimedia speakers' content and upload it to Commons using OAuth and the bots applicable as they do now for their own content. But for wikimedians strictly, yes, they could do with an in-built record wizard on the projects to decrease the reliance on bots to upload 100% of recordings, and the developer maintaining them and the record wizard. Eavqwiki (talk) 11:05, 22 November 2019 (UTC)[reply]

Seems like some overlap/redundancy with my proposal here: Community Wishlist Survey 2020/Wiktionary/Fix the AddAudio script to upload audio to Commons seamlessly. I don't have strong feelings on which tool/script is the best one. —Justin (koavf)TCM 21:00, 25 November 2019 (UTC)[reply]