[00:10:29] welcome Sergio [05:13:24] Note that Google AI is confused about how to subtract 10 hours from a time! Over the last month or so I've gradually put together the equivalent in Wikidata, so now for any QID and date, you can get the time of sunset. Even with a few approximations, it's correct within a minute! (The answer is in UTC because we still haven't got timezones.) This is my candidate for a [05:13:24] function you could copy onto any QID on any date and the time output could be configured into any language. : https://tools-static.wmflabs.org/bridgebot/4b4e5bcf/file_73668.jpg [05:13:51] Z28005 @vrandecic (re @u99of9: Note that Google AI is confused about how to subtract 10 hours from a time! Over the last month or so I've gradually put togeth...) [05:18:12] I'm still planning to clear up some of those approximations (e.g. Z27475 is pretty rough), so the accuracy should further improve. (re @u99of9: Note that Google AI is confused about how to subtract 10 hours from a time! Over the last month or so I've gradually put togeth...) [07:18:23] I think Z24392 used to work. Although there's more info than previously, there's still not quite enough to immediately track down where the BigInt conversion issue is. [17:02:33] Friendly reminder that our meeting is in 30 minutes. See you there! (re @Sannita: Hi all! Our next Volunteers' Corner will be held on Monday, September 8, at 17:30 UTC. [17:02:34] If you have questions or ideas to discus...) [17:20:03] hm, how functions can be used in items?.. (re @u99of9: Note that Google AI is confused about how to subtract 10 hours from a time! Over the last month or so I've gradually put togeth...) [22:40:09] See https://www.wikifunctions.org/wiki/Wikifunctions:Embedded_function_calls. At this stage it's only available on some wikis. (re @AntonVChupin: hm, how functions can be used in items?..) [22:53:08] From what happened on recentchanges, I assume the VC discussed the need for Z22074 and T383842? Was any guidance or commitment given on this? At T390226 it was listed as not a high priority, but I have long held that functions like Z28016 are not possible without it. (Instead in the function signatures the QIDs would need to be shoved into a list which has numerous [22:53:08] downsides, inc [22:53:09] luding not being able to label which element means what.) (re @Sannita: Friendly reminder that our meeting is in 30 minutes. See you there!) [22:59:20] @dvd_ccc27919 your attempt at Z28035 is impressively able to pass the one test. Can you say more on why it doesn't "really" work? [23:50:36] PS another attempt was here Z21216, which others have tried to use for a similar language fragment function Z27174. (re @u99of9: @dvd_ccc27919 your attempt at Z28035 is impressively able to pass the one test. Can you say more on why it doesn't "really" work...)