[05:59:48] I've disconnected Z23766 because although all three implementations fail, one counts as passing (why?) which tricks the bot into promoting that implementation. [09:51:43] I’ve disconnected Z30487, so the successful evaluations of Z23766 now pass. The Python implementation also shows as passing although it times out, but I guess that’s *T373607*. I’ve added Z35195 to Z23120 as an equivalent test, and that times out and shows as failing ✅. The Python timeout should be investigated. (re @u99of9: I've disconnected Z23766 because although all t [09:51:43] [09:51:44] hree implementations fail, one counts as passing (why?) which tricks the bot into...) [10:41:27] *Abstract Wikipedia Newsletter #248 [10:41:27] ** A higher meaning [10:41:29] * Recent Changes in the software [10:41:30] * News in Types: way forward for natural language generation types [10:41:32] * Recording of the May 11 Volunteers' Corner [10:41:33] * Fresh Function weekly: 40 new Functions [10:41:35] https://www.wikifunctions.org/wiki/Wikifunctions:Status_updates/2026-05-15 [10:41:41] I'd be particular curious about the question in the News in Types section [10:42:40] I also haven't been able to form a clear opinion. [10:44:58] If they are all clarified enough to mean they won't later need changes, then yes, releasing them sounds fair. [10:51:46] I will be asking questions about the places that are not clear in the next few days. But indeed, I am not sure whether these types would be surviving the test of times. [13:26:30] A question: do we agree in the use of "fake lexemes" and "fake items"? Since they are a fundamental part of Semantic Unit and Syntactic Unit current definition (re @vrandecic: Abstract Wikipedia Newsletter #248 [13:26:30] * A higher meaning [13:26:32] * Recent Changes in the software [13:26:33] * News in Types: way forward for natural ...) [13:35:47] This, really. As long as it's good enough such that the risk that we need to erase the work of contributors or add a very large maintenance task is negligible, I think we should just go ahead and add them. If this isn't guaranteed then I can't support it. (re @u99of9: If they are all clarified enough to mean they won't later need changes, then yes, releasing them [13:35:47] sounds fair.) [13:36:17] I'm not a fan of these, honestly. If you need it, add it on Wikidata. (re @dvd_ccc27919: A question: do we agree in the use of "fake lexemes" and "fake items"? Since they are a fundamental part of Semantic Unit and Sy...) [13:40:42] Fake Lexemes would be needed to encode mainly places names and people's name. We are talking potentially of an ever growing set of millions or billions of needed lexemes per language. [13:40:42] Fake items for now are not necessary, but eventually could be used to describe the specific things that we are talking about on Abstract Wikipedia (re @Feeglgeef: I'm not a fan of these. If you need it, add it on Wikidata.) [13:45:31] I am not a huge fan of the term "fake Lexeme", would prefer something like "transient Lexemes" or "on-the-fly Lexemes" or "non-permanent Lexemes" (same for Items). I am sure, there will be even better proposals. [16:16:05] +1, maybe "local lexeme" ? (re @vrandecic: I am not a huge fan of the term "fake Lexeme", would prefer something like "transient Lexemes" or "on-the-fly Lexemes" or "non-p...) [16:16:53] -> locseme (re @NicolasVIGNERON: +1, maybe "local lexeme" ?)