[03:57:26] Thanks for trying this out at Z29333 @dvd_ccc27919. I've fixed it and it works, but unfortunately I had to write a helper function Z29334 (silly lossy wrapper of your well-specified Z22997) to ensure that the function only had two arguments (and the second was a list). This is the first of the proliferation of poorly-specified duplicate functions that I foreshadowed [03:57:26] in T383842. [03:57:27] IMO this is even stronger evidence that T390226 should be part of essential work, not "medium" priority. @vrandecic can I ask for your involvement here, or how should I communicate this need to "product"? (re @u99of9: I made this: Z29324 It's a bit technical, but I think it will help construct the inner bit of these Wiktionary tables.) [08:51:00] Actually Z22074 somehow currently works (re @u99of9: Thanks for trying this out at Z29333 @dvd_ccc27919. I've fixed your test and it works, but unfortunately I had to write a helper...) [09:00:43] Interesting, but only for one of two tests, so I can't rely on it in compositions. (re @dvd_ccc27919: Actually Z22074 somehow currently works) [09:03:51] Actually, even the one that is marked as passing doesn't work in "try a function" (re @u99of9: Interesting, but only for one of two tests, so I can't rely on it in compositions.) [09:21:59] It returns a function call. Wrapped with Echo, it can evaluate the function call. (re @u99of9: Actually, even the one that is marked as passing doesn't work in "try a function") [09:23:36] So should the real apply3 compose echo(Z22074())? (re @Al: It returns a function call. Wrapped with Echo, it can evaluate the function call.) [09:25:37] I’m just looking into it, but I have a lot on this weekend… (re @u99of9: So should the real apply3 compose echo(Z22074())?) [09:27:48] No rush, this has been a very longstanding deficit. (re @Al: I’m just looking into it, but I have a lot on this weekend…) [09:29:17] It feels wrong to rely on a code implementation for something so list dependent. [09:41:19] Compare and contrast the implementations for Z28016 🤷‍♂️ (re @u99of9: It feels wrong to rely on a code implementation for something so list dependent.) [10:11:45] Yes, Z28032 seems simple enough. That might work. (re @Al: Compare and contrast the implementations for Z28016 🤷‍♂️) [12:48:54] oh, I disconnected the Swedish implementation today because the only test failed. [12:48:54] @Jan_ainali I suggested a solution in the talk page. [12:48:55] I don't have access to a laptop browser at the moment. (re @Al: Yes, that’s an odd one. It seems to be working correctly for Breton, apart from in the test. We shouldn’t have a function with n...) [12:50:15] I don't know which implementation you are referring to here. (re @Npriskorn: oh, I disconnected the Swedish implementation today because the only test failed. [12:50:16] @Jan_ainali I suggested a solution in the talk...) [12:59:06] Ah, I now see. And that it already has been reverted as it seems all good. (re @Jan_ainali: I don't know which implementation you are referring to here.) [14:35:14] 🤔 All six tests pass for Z29349 but I’m leaving it disconnected until I have time to validate the impact, because it looks like it would become the preferred implementation. (re @u99of9: Yes, Z28032 seems simple enough. That might work.) [14:52:33] 5956 [15:31:49] sorry for not including a link. (re @Jan_ainali: I don't know which implementation you are referring to here.) [17:17:22] I also implemented Z29355 for Z21132, which is similar to Z22074, except it takes a string ZID for the function. It also returns a function call rather than an evaluation, and this evaluates when wrapped by Z801. (re @Al: 🤔 All six tests pass for Z29349 but I’m leaving it disconnected until I have time to validate the impact, because it looks like ...) [21:35:23] Let's give it a go. (re @Al: 🤔 All six tests pass for Z29349 but I’m leaving it disconnected until I have time to validate the impact, because it looks like ...)