[10:36:18] Hallo [10:37:24] Welcome to the new episode of the never-ending series Amir Aharoni's pedantic questions about translatable messages in the WikiLambda extension! [10:38:19] A few recently added messages use the words "function", "ZFunction", and "Wikifunction". [10:38:30] I can translate them all, but I nevertheless wonder: [10:38:59] Are they all the same thing, or is there any difference? [10:40:07] The Glossary doesn't have an entry for "Wikifunction". [10:44:39] as I understand it ZFunctions are the REPESENTATION of a function in json, wikifunctions is the collection of the Zfunctions [11:10:17] Wikifunctions, with an s in the end, is the site. [11:10:31] Or the plural form of "Wikifunction". [11:10:56] But the singular form "Wikifunction" doesn't appear in the Glossary. [11:14:49] which message has “Wikifunction” without the s? I can’t find it [11:45:55] There are none, but I thought that some of them may be the plural form of "Wikifunction". Am I overthinking? :) [11:46:02] Does it always refer to the site? [11:47:25] E.g., Wikibase messages usually don't say "Wikidata", but use a magic word to refer to the repo name. [11:52:06] *looks at en.json again* [11:52:46] “Pages with Wikifunctions calls”, hmm [11:53:08] how would that sound if it used {{SITENAME}}? “Pages with Not Wikilambda calls” [11:53:19] I feel like that should maybe just be “function calls” [11:54:02] (some other messages already use “function call” anyways) [11:58:21] Not sure. Maybe Real Designers will have a better idea :) [12:00:57] I vote for function calls :) [14:10:14] We'll certainly need a pass over our messages and be more consistent in the use of the glossary [17:45:42] Where should I look for the definitive discussion about Wikidata Q-identifiers? I understand some of the arguments on each side, of natural language vs. stable machine IDs, but I’m curious to know if there has been any retrospective about the decisions that were made. [18:34:27] Asking because the Z-ids will have most of the same strengths and weaknesses, and because I just tuned into a conversation which I need a decoder ring to approach 😉 (re @Adam: Where should I look for the definitive discussion about Wikidata Q-identifiers? I understand some of the arguments on each side, of natural language vs. stable machine IDs, but I’m curious to know if there has been any retrospective about the decision [20:04:53] Hi Adam! I'm not sure what you mean with "definite discussion about Q-Identifiers"? [22:44:39] https://github.com/bnjbvr/rouille is apparently trending on GitHub. (In addition to a French Rust, it links to similar efforts in Dutch and German.) [23:19:48] I disagree with the German one – they’ve translated “let” as „wenn“ when, going by the Germans math proofs I’ve seen, it should probably be „sei“ [23:20:44] the conjunctive case can make for wonderfully concise sentences: „Sei n ∈ ℕ, x, y ∈ ℝ, p prim. Dann ist…” [23:22:11] AFAIK English proofs usually need two words there, “let X be…” [23:22:32] I disagree with the German one – they’ve translated “let” as „lass“ when, going by the Germans math proofs I’ve seen, it should probably be „sei“ [23:23:24] yeah, "let x be" is what I'm familiar with