[00:58:24] I'm having trouble constructing wikidata statements for validation at Z29836. I think the difference might be in the highlighted sections. : https://tools-static.wmflabs.org/bridgebot/865f49e4/file_75711.jpg [01:06:42] I think you’d fail with references too. Have you tried using Z23723? [01:07:23] I just looked that up actually, thanks, I'll try. (re @Al: I think you’d fail with references too. Have you tried using Z23723?) [01:13:44] I'm confused at why subject entity reference (item) and subject entity reference (lexeme) arguments expect lists (re @wikilinksbot: Z23723 – emulate Wikidata statement object) [01:14:34] Just so that they are optional. (re @u99of9: I'm confused at why subject entity reference (item) and subject entity reference (lexeme) arguments expect lists) [01:17:17] I've made the change, but I'm still seeing non-matching (different) output (re @u99of9: I just looked that up actually, thanks, I'll try.) [01:19:35] Are you using nullary void for Z23723K9? (re @u99of9: I've made the change, but I'm still seeing non-matching (different) output) [01:20:25] No, I ignored that. I take it I should? (re @Al: Are you using nullary void for Z23723K9?) [01:21:26] Should it be a single item list with a void item, or the whole argument should be the void? (re @u99of9: No, I ignored that. I take it I should?) [01:22:45] The whole argument should be void, IIRC [01:23:33] I've put that in now, but I think I'm actually getting a *lot* of differences: Expected result: ["Z6003" {"Z1K1" "Z7" "Z7K1" "Z23723" "Z23723K1" ["Z6091" {"Z1K1" "Z6091" "Z6091K1" "Q3224"}] "Z23723K2" ["Z6095"] "Z23723K3" {"Z1K1" "Z6092" "Z6092K1" "P421"} "Z23723K4" ["Z6091" {"Z1K1" "Z6091" "Z6091K1" "Q7056"}] "Z23723K5" ["Z6095"] "Z23723K6" ["Z1"] "Z23723K7" "Z6042""Z [01:23:33] 23723K8":[ [01:23:34] "Z6007" {"Z1K1" "Z6007" "Z6007K1" {"Z1K1" "Z6092" "Z6092K1" "P1264"} "Z6007K2" {"Z1K1" "Z6091" "Z6091K1" "Q1777301"} "Z6007K3" "Z6021"}] "Z23723K9" {"Z1K1" "Z7" "Z7K1" "Z26199"} "Z23723K10" "Z6021"} {"Z1K1" "Z7" "Z7K1" "Z23723" "Z23723K1" ["Z6091" {"Z1K1" "Z6091" "Z6091K1" "Q3224"}] "Z23723K2" ["Z6095"] "Z23723K3" {"Z1K1" "Z6092" "Z6092K1" "P421"} "Z23723K4" ["Z6091"{" [01:23:34] Z1K1":"Z60 [01:23:36] 91","Z6091K1":"Q7069"}],"Z23723K5":["Z6095"],"Z23723K6":["Z1"],"Z23723K7":"Z6042","Z23723K8":["Z6007",{"Z1K1":"Z6007","Z6007K1":{"Z1K1":"Z6092","Z6092K1":"P1264"},"Z6007K2":{"Z1K1":"Z6091","Z6091K1":"Q36669"},"Z6007K3":"Z6021"}],"Z23723K9":{"Z1K1":"Z7","Z7K1":"Z26199"},"Z23723K10":"Z6021"},{"Z1K1":"Z7","Z7K1":"Z23723","Z23723K1":["Z6091",{"Z1K1":"Z6091","Z6091K1":"Q322 [01:23:36] 4"}],"Z2372 [01:23:37] 3K2":["Z6095"],"Z23723K3":{"Z1K1":"Z6092","Z6092K1":"P421"},"Z23723K4":["Z6091",{"Z1K1":"Z6091","Z6091K1":"Q4823536"}],"Z23723K5":["Z6095"],"Z23723K6":["Z1"],"Z23723K7":"Z6042","Z23723K8":["Z6007"],"Z23723K9":{"Z1K1":"Z7","Z7K1":"Z26199"},"Z23723K10":"Z6021"}] [01:23:39] Actual result: ["Z6003" {"Z1K1" "Z6003" "Z6003K1" {"Z1K1" "Z6091" "Z6091K1" "Q3224"} "Z6003K2" {"Z1K1" "Z6092" "Z6092K1" "P421"} "Z6003K3" {"Z1K1" "Z6091" "Z6091K1" "Q7056"} "Z6003K4" {"Z1K1" "Z6040" "Z6040K1" "Z6042"} "Z6003K5" ["Z6007" {"Z1K1" "Z6007" "Z6007K1" {"Z1K1" "Z6092" "Z6092K1" "P1264"} "Z6007K2" {"Z1K1" "Z6091" "Z6091K1" "Q1777301"} "Z6007K3" {"Z1K1" [01:23:39] "Z6020","Z6020K1" [01:23:40] :"Z6021"}}],"Z6003K7":{"Z1K1":"Z6020","Z6020K1":"Z6021"}},{"Z1K1":"Z6003","Z6003K1":{"Z1K1":"Z6091","Z6091K1":"Q3224"},"Z6003K2":{"Z1K1":"Z6092","Z6092K1":"P421"},"Z6003K3":{"Z1K1":"Z6091","Z6091K1":"Q7069"},"Z6003K4":{"Z1K1":"Z6040","Z6040K1":"Z6042"},"Z6003K5":["Z6007",{"Z1K1":"Z6007","Z6007K1":{"Z1K1":"Z6092","Z6092K1":"P1264"},"Z6007K2":{"Z1K1":"Z6091","Z6091K1":"Q [01:23:41] 36669"},"Z6 [01:23:43] 007K3":{"Z1K1":"Z6020","Z6020K1":"Z6021"}}],"Z6003K7":{"Z1K1":"Z6020","Z6020K1":"Z6021"}},{"Z1K1":"Z6003","Z6003K1":{"Z1K1":"Z6091","Z6091K1":"Q3224"},"Z6003K2":{"Z1K1":"Z6092","Z6092K1":"P421"},"Z6003K3":{"Z1K1":"Z6091","Z6091K1":"Q4823536"},"Z6003K4":{"Z1K1":"Z6040","Z6040K1":"Z6042"},"Z6003K7":{"Z1K1":"Z6020","Z6020K1":"Z6021"}}] [01:31:15] It’s hard to tell, because the function calls aren’t evaluated for the expected result display 🤷‍♂️ (re @u99of9: I've put that in now, but I think I'm actually getting a *lot* of differences: Expected result: ["Z6003",{"Z1K1":"Z7","Z7K1":"Z2...) [01:31:27] Here's a simpler version of the test at emulate: Z29846 [01:34:46] I fixed that for you. (re @u99of9: Here's a simpler version of the test at emulate: Z29846) [01:37:48] Ah, so it's probably the qualifiers that also have to be nullary void [01:38:43] Yes, if there are none, the key is omitted rather than being an empty list. (re @u99of9: Ah, so it's probably the qualifiers that also have to be nullary void) [01:39:07] Yes! This works now: Z29836. I'll tweak the argument titles to help me next time. (re @u99of9: Ah, so it's probably the qualifiers that also have to be nullary void) [01:44:28] It’s actually “(nullary void if omitted)”. If you want an empty list, you supply an empty list. (re @u99of9: Yes! This works now: Z29836. I'll tweak the argument titles to help me next time.) [01:45:19] There's something strange going on here. I've tried to add a title a few times in a row, and the edits go through to the item in some kind of list, but don't show up here. : https://tools-static.wmflabs.org/bridgebot/9f078403/file_75712.jpg [01:47:30] That’s *T405839* (re @u99of9: There's something strange going on here. I've tried to add a title a few times in a row, and the edits go through to the item in...) [01:50:30] Okay great - I won't try to manually fix it! (re @Al: That’s T405839) [01:57:19] The first item inheritance chain turns out not terrible: Z29847. (But sometimes the first item will hit a dead end?) (re @Al: It’s the nested indirection that troubles me. Perhaps we have a shortlist enumeration of countries with more than one time zone,...) [02:02:37] After learning how to get around this, I still think it would be better if we could construct a statement in the UI. Then the test could look like the first figure instead of the second : https://tools-static.wmflabs.org/bridgebot/2e305a86/file_75713.jpg [02:02:39] https://tools-static.wmflabs.org/bridgebot/949c1f15/file_75714.jpg [02:03:55] David or Daphne maybe this is for one of you? (re @u99of9: After learning how to get around this, I still think it would be better if we could construct a statement in the UI. Then the te...) [02:09:39] I doubt you’d often reach a dead end, and the data should probably be fixed if you do. (re @u99of9: The first item inheritance chain turns out not terrible: Z29847. (But sometimes the first item will hit a dead end?)) [02:13:05] Some things span timezones. Maybe not always fixable. https://www.wikifunctions.org/view/en/Z29835?call=%7B%22Z1K1%22%3A%22Z7%22%2C%22Z7K1%22%3A%22Z29835%22%2C%22Z29835K1%22%3A%7B%22Z1K1%22%3A%22Z7%22%2C%22Z7K1%22%3A%22Z6821%22%2C%22Z6821K1%22%3A%7B%22Z1K1%22%3A%22Z6091%22%2C%22Z6091K1%22%3A%22Q186545%22%7D%7D%7D (re @Al: I doubt you’d often reach a dead end, and the [02:13:06] data shoul [02:13:06] d probably be fixed if you do.) [03:14:50] Hi hi! Is there similar concept of sandbox in Wikifunctions? I found these sandbox functions: Z24074, Z24093, Z10119, Z15104, and Z26295, can we use them for our sandbox to create implementations and/or tests? Or is there better approach? [03:14:51] For context, this is for an event to introduce Wikifunctions [03:20:48] Yes, feel free to use them as you wish, remembering that others may also change them. Nothing else relies on them. (re @nikolaskhf: Hi hi! Is there similar concept of sandbox in Wikifunctions? I found these sandbox functions: Z24074, Z24093, Z10119, Z15104, an...) [03:29:18] Thank you! (re @u99of9: Yes, feel free to use them as you wish, remembering that others may also change them. Nothing else relies on them.) [03:43:10] We will need a function where we can feed in a location QID and a date, and determine whether daylight savings time applies. https://en.wikipedia.org/wiki/Daylight_saving_time_by_country [03:47:31] @dvd_ccc27919 this looks like it has overlap with your work on seasonal holidays etc. (re @u99of9: We will need a function where we can feed in a location QID and a date, and determine whether daylight savings time applies. htt...) [07:07:53] I have a patch up to fix this (re @u99of9: There's something strange going on here. I've tried to add a title a few times in a row, and the edits go through to the item in...) [07:11:16] I think David mentioned to me when working on view statements, that its still possible to create statements using Wikidata statement. But I am perhaps misunderstanding the request (re @u99of9: After learning how to get around this, I still think it would be better if we could construct a statement in the UI. Then the te...) [07:40:20] I could create statements, but they didn't match the format of the fetched ones. (re @Daphne: I think David mentioned to me when working on view statements, that its still possible to create statements using Wikidata stat...) [07:55:18] I think David M is enjoying thanksgiving. But I’ll check the examples and write up something to check after [08:03:27] The problem is that Z6003K5 and Z6003K6 can be omitted from an object created by the fetch, but not when constructed in the UI. Happy Thanksgiving! (re @Daphne: I think David M is enjoying thanksgiving. But I’ll check the examples and write up something to check after) [08:24:29] The highlighted differences don't matter? Are they just a reference difference, or a UI display difference? (re @u99of9: I'm having trouble constructing wikidata statements for validation at Z29836. I think the difference might be in the highlighted...) [08:48:45] I think they are just different representations of the same thing. (re @u99of9: The highlighted differences don't matter? Are they just a reference difference, or a UI display difference?) [08:51:11] We should pick one and stick to it. (re @Al: I think they are just different representations of the same thing.) [10:45:54] …or we should consistently have both. A reference should display as a simple link that is expandable to show its de-referenced form. I think this only happens for Z41 and Z42, currently, but it’s consistent with the behaviour when there is a display function. (re @u99of9: We should pick one and stick to it.) [10:54:53] Yes I'm happy with the way that display function outputs can be deconstructed. Booleans are fine. This sounds good. (re @Al: …or we should consistently have both. A reference should display as a simple link that is expandable to show its de-referenced f...) [13:12:43] I agree that we should prefer lexemes over labels. I’m waiting for Z6820 to come back before working on this, but there are other fallbacks like P1476, P1448, P1559 and P1705 that we should consider. We might also look at P5972, particularly for the case where the P5137 is missing for the target language but present in a different language. (re @vrandecic: I really [13:12:43] enjoyed writ [13:12:43] ing the "Finishing each other's sentences section", and would love to cooperate in this direction with folk...) [13:28:38] Would it ever be the case that if there is a P5972*//P5137 chain, there shouldn't be a direct P5137 as well? [14:22:20] Ah, that I cannot say. Certainly it is the case that the number of translations varies when the same item is linked. (re @vrandecic: Would it ever be the case that if there is a P5972*//P5137 chain, there shouldn't be a direct P5137 as well?) [14:53:00] A partial example is L1330150-S1. This does not link to Q15978631 (only to Q5) but is present as a translation for L3080-S1, which has links to Q5, Q15978631 and Q3238275. (re @Al: Ah, that I cannot say. Certainly it is the case that the number of translations varies when the same item is linked.) [16:38:58] Deos someone here know why Z29689 failed in a few tests? [16:38:58] cache related? [19:04:14] Yes. All clear now 👍 (re @Npriskorn: Deos someone here know why Z29689 fails in a few tests? [19:04:15] cache related?)