[00:39:46] Why don't functions returning quantities work when embedded? I've tried the simple read function: {{#function:Z25785|20.0 mi}} and it displays "content error". [02:39:41] I thought maybe the entries would be randomized but if we both got "Wikipedia Babel" (one of many choices) then that's a coincidence or it's not random (re @waldyrious: Is there any explicit rule to determine which entries get promoted to the main list vs the More proposals page? E.g. currently "...) [02:42:59] We should write a page called something like "Policies for Tests", then when this page comes up, we could link to it: : https://tools-static.wmflabs.org/bridgebot/9444ac2f/file_74783.jpg [04:18:31] It’s basically working (with the visual editor) but the specific value fails in preview, presumably with a cached error. (re @u99of9: Why don't functions returning quantities work when embedded? I've tried the simple read function: {{#function:Z25785|20.0 mi}} ...) [05:03:22] Hmm, good to know. I haven't tried with visual editor before! (re @Al: It’s basically working (with the visual editor) but the specific value fails in preview, presumably with a cached error.) [05:13:20] De we have a help page for tests? We should link to a page like that too from there. (re @u99of9: We should write a page called something like "Policies for Tests", then when this page comes up, we could link to it:) [06:39:47] I just tried ChatGPT on sunset, and it did well: : https://tools-static.wmflabs.org/bridgebot/fbdeb426/file_74785.jpg [06:42:54] ChatGPT does well at this conversion, but oddly has the exact same significant figure problem (rounding to 3sf instead of 4sf) as I used to. Is that suspicious? : https://tools-static.wmflabs.org/bridgebot/54cff1c1/file_74786.jpg [06:47:57] ChatGPT goes horribly wrong with the calculations : https://tools-static.wmflabs.org/bridgebot/84cf465f/file_74787.jpg [08:38:11] That’s not the exact conversion from kg to oz. 1 kg = 35.2739619495804 oz, which gives 7.75921341. I asked it how to round this and it said: [08:38:12] Metrology and physics usually follow a simple hierarchy of rounding discipline: [08:38:13] 1. *Match significant figures to the least precise input.* [08:38:15] Your _0.001234 eV/c²_ has four significant figures, so the converted result should too: *7.759 × 10⁻³⁸ oz* or *7.760 × 10⁻³⁸ oz*. [08:38:16] 2. *Don’t add precision you didn’t start with.* [08:38:18] It’s tempting to carry every decimal, but extra digits imply confidence you don’t have. [08:38:19] 3. *Round last, not mid-calculation.* [08:38:21] Keep internal precision high until the final step, then round once. [08:38:22] 4. *If it’s for display, not calculation, favour clarity.* [08:38:24] A clean figure like *7.76 × 10⁻³⁸ oz* reads naturally and still respects the data. [08:38:25] In short: preserve precision internally, express significance externally. (re @u99of9: ChatGPT does well at this conversion, but oddly has the exact same significant figure problem (rounding to 3sf instead of 4sf) a...) [08:39:59] Usually they are the more supported, but the threshold is not fixed in stone (re @waldyrious: Is there any explicit rule to determine which entries get promoted to the main list vs the More proposals page? E.g. currently "...) [08:40:44] We can decide it's the first 8? 10? 5? [08:40:51] I would go with 8 [08:41:19] And again I have connection issues at home, I really don't know why I'm spending all this money for [08:41:31] So I will update the list when I can [09:18:38] Five would be too few and thirteen would be too many, unless there are more than six ties for sixth place 🤷‍♂️ (re @Sannita: We can decide it's the first 8? 10? 5?) [09:38:29] Why not 10? I'm happy to bend the decimal hegemony towards a binary/octal option, but it might be intriguing choice to people. (I guess it would make sense if we were planning the next round to be a tournament-style series of pairwise matches, but I don't think that's the case 🙂) (re @Sannita: I would go with 8) [10:02:59] There is. It is called "opinion/bias" presented as "judgment" or "hierarchy". (re @waldyrious: Is there any explicit rule to determine which entries get promoted to the main list vs the More proposals page? E.g. currently "...) [13:37:55] I have been going for "less than 20" [13:38:37] Right now it is just a bit over that, and I was thinking of relegating out the ones with 3 supports, and raise the threshold to 4 support [13:38:56] If you think that's too many, we can reduce that number [13:40:03] I am randomizing them manually about once a day. If there is a way to randomize it automatically per view, that would be awesome! (re @harej: I thought maybe the entries would be randomized but if we both got "Wikipedia Babel" (one of many choices) then that's a coincid...) [16:23:45] @u99of9 I tried switching to your float implementation for the subnormals. Looks good so far, some tests are stuck in cache for now [16:24:32] 1986 [16:35:25] I think someone implemented pseudo-random number generation with parser functions once [16:35:49] is wikifunctions capable of pseudo random number generation? [16:35:57] I also had that memory, but I cannot find that [16:36:10] No, Wikifunctions is intenionally functional [16:38:02] Pseudo random number generation is functional isn't it? It's simulated entropy based on a function or program that is run [16:38:08] Or is there another distinction I am missing [16:38:22] I think I have deleted some randomness functions quite some time ago. [16:38:39] I mean, the results are deterministic with regards to inputs [16:39:10] Expectation of deterministic output. Got it [16:39:40] I guess you could put in a seed as the input, but it would result in the same exact randomness string every time. Then you'd have to randomize the input