[01:25:19] This only happened once, but I don't think I've seen it before: : https://tools-static.wmflabs.org/bridgebot/10e9a343/file_69927.jpg [03:34:35] How do I test a result against a persistent object value? This currently has an expected result equal to the reference. Z23904 [03:36:30] That's not the issue, the issue is with the actual result or equality function (re @u99of9: How do I test a result against a persistent object value? This currently has an expected result equal to the reference not the v...) [03:37:12] Expected result as shown doesn't matter, it's just what you are plugging into the equality function [03:38:12] It's true that the actual result is off by a rounding error in the significand of 1. Maybe the issue I have is with the way the failure is displayed? : https://tools-static.wmflabs.org/bridgebot/b7c8eafd/file_69928.jpg [03:39:01] I disagree with the changing that (re @u99of9: It's true that the actual result is off by a rounding error in the significand of 1. Maybe the issue I have is with the way the ...) [03:39:28] Expected result can never be objective, that's how equality functions are designed [03:40:44] Okay, I suppose I just have to get used to it, and I have to look up the value I need to match to see how far off I am. [03:40:52] Yeah (re @u99of9: Okay, I suppose I just have to get used to it, and I have to look up the value I need to match to see how far off I am.) [03:42:00] I suppose you could dereference actually? (re @Feeglgeef: I disagree with the changing that) [03:48:24] Yes, I think I'd prefer that - it would make it more comparable to regular test failure reports. (re @Feeglgeef: I suppose you could dereference actually?) [03:55:57] I'm currently getting this at Z23902, and there's not even a carrot to expand to circumvent the display function and see the raw keys. : https://tools-static.wmflabs.org/bridgebot/0f1dc53b/file_69929.jpg [05:49:12] I wonder if there is a code conversion issue that prevents a float64 value even getting to the display function? (re @u99of9: I'm currently getting this at Z23902, and there's not even a carrot to expand to circumvent the display function and see the raw...) [06:49:01] Yes, I’m getting that all the time with that function (unless value is 0) 🤷‍♂️ (re @u99of9: I'm currently getting this at Z23902, and there's not even a carrot to expand to circumvent the display function and see the raw...) [07:01:53] I get exactly the same message for English with the value 3.141592653589793 on Z21925 (referencing Z21956) (re @u99of9: I'm currently getting this at Z23902, and there's not even a carrot to expand to circumvent the display function and see the raw...) [07:04:57] …and in Z21642 (re @u99of9: I'm currently getting this at Z23902, and there's not even a carrot to expand to circumvent the display function and see the raw...) [08:38:54] I think I finally managed to kick the error out of the cache by making this test Z23912. Maybe they're all working again? (re @Al: I get exactly the same message for English with the value 3.141592653589793 on Z21925 (referencing Z21956)) [08:58:51] Looks like it, thank you… surely a case for *T391319* (re @u99of9: I think I finally managed to kick the error out of the cache by making this test Z23912. Maybe they're all working again?) [09:56:13] I noticed that Z23905 gives the significand 2570638124657943 for final term == 6 (timing out for 7). There were also differences of 1 [[https://www.wikifunctions.org/wiki/Z18454?uselang=en&oldid=178801|here]] and [[https://www.wikifunctions.org/wiki/Z22196?uselang=en&oldid=178803|here]]. Do we need a less precise comparison function? (re @u99of9: It's true that the actual [09:56:13] result [09:56:13] is off by a rounding error in the significand of 1. Maybe the issue I have is with the way the ...) [10:00:43] Mentioned at T391435#10742767 (re @u99of9: This only happened once, but I don't think I've seen it before:) [10:06:09] Mentioned at T391319#10742743. We probably need a new ticket for the improper error message. (re @Al: Looks like it, thank you… surely a case for T391319) [10:11:38] really? I get this. When I was off by 1 before, I rewrote the composition in a simpler way to reduce the number of round off opportunities, which I thought had fixed it for all terms >= 1. : https://tools-static.wmflabs.org/bridgebot/7880c8f3/file_69931.jpg [10:13:51] But in any case, I agree it will often help us to also have a less demanding comparator. I just haven't jumped in yet. (re @Al: I noticed that Z23905 gives the significand 2570638124657943 for final term == 6 (timing out for 7). There were also differences...) [10:14:43] Yes, really 🤔 (re @u99of9: really? I get this. When I was off by 1 before, I rewrote the composition in a simpler way to reduce the number of round off op...) [10:15:55] Yes, really 🤔 : https://tools-static.wmflabs.org/bridgebot/b653e04a/file_69932.jpg [10:17:07] I'm not sure if anything in particular was improper. If a page fails to load, it's good to get some kind of message. As long as things don't fail often I don't mind refreshing once in a blue moon. But I posted just in case others had seen an increased frequency recently (or if it was related to the python list problem). (re @Al: Mentioned at T391319#10742743. We [10:17:07] probably need a n [10:17:07] ew ticket for the improper error message.) [10:17:59] Yeah, that’s why I mentioned it there 😎 (re @u99of9: I'm not sure if anything in particular was improper. If a page fails to load, it's good to get some kind of message. As long as ...) [10:18:19] This is extremely odd... everything here is deterministic!? (re @Al: Yes, really 🤔) [10:20:46] I'm at the URL: https://www.wikifunctions.org/view/en/Z23902 I type 6, run, then expand the result... (re @u99of9: This is extremely odd... everything here is deterministic!?) [10:22:17] Ah, I’m at Z23905 (re @u99of9: This is extremely odd... everything here is deterministic!?) [10:22:27] ohhh, but when I run try this composition it gives yours. OK, so it's from teh other implemen... [yes] (re @u99of9: I'm at the URL: https://www.wikifunctions.org/view/en/Z23902 I type 6, run, then expand the result...) [10:23:02] Yes 😎 (re @u99of9: ohhh, but when I run try this composition it gives yours. OK, so it's from teh other implemen... [yes]) [10:27:13] Now we have even more incentive I guess Z23916 (re @u99of9: But in any case, I agree it will often help us to also have a less demanding comparator. I just haven't jumped in yet.) [10:31:09] I get queasy just thinking about it… 😂 (re @u99of9: Now we have even more incentive I guess Z23916) [10:40:24] I'll get to it at some point, but I really want subnormals dealt with properly first. (re @Al: I get queasy just thinking about it… 😂) [10:42:36] I miss Denny! (re @u99of9: @vrandecic this is probably for you.) [11:31:38] Denny is really busy this month, sorry guys [11:32:05] we're working for you though, this is why he's not answering messages here [11:38:23] Wikifunctions will be deployed on the Dagbani Wikipedia on April 15! [11:38:24] see https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/NJLULEJ5LJMATJSOPJRA3U5N5WAAEUFB/ [11:38:25] I can't wait to see it! (and I'm wondering where to learn more and follow this) [11:39:15] We understand… but we still miss him! 😎 (re @Sannita: we're working for you though, this is why he's not answering messages here) [11:39:26] (Actually, there might be a slight delay in deployment, we'll keep you posted about this 😞) (re @Nicolas: Wikifunctions will be deployed on the Dagbani Wikipedia on April 15! [11:39:27] see https://lists.wikimedia.org/hyperkitty/list/wikimedia-l...) [11:41:26] (it's a very hard time to be a community liaison 😂) [11:48:04] @Sannita The above link does not work, is it just me? [11:48:56] @bunnypranav you mean the wikimedia-l link? works for me [11:49:06] Works for me 🤷‍♂️ (re @wmtelegram_bot: @Sannita The above link does not work, is it just me?) [11:50:06] Oh, @Sannita it was this link in irc. : https://tools-static.wmflabs.org/bridgebot/67532721/file_69933.jpg [11:51:33] this link in irc, seems to be cut tho. : https://tools-static.wmflabs.org/bridgebot/9b6e3592/file_69934.jpg [11:52:04] I see, it doesn't work because it's a reply (re @bunnypranav: this link in irc, seems to be cut tho.) [11:52:13] full link is https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/NJLULEJ5LJMATJSOPJRA3U5N5WAAEUFB/ [11:52:33] Ah, the split irc msg confused me, thought you posted it. [11:52:45] that does indeed work (re @Sannita: full link is https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/NJLULEJ5LJMATJSOPJRA3U5N5WAAEUF...) [11:52:55] no, it was Nicolas who did [11:53:27] realized a bit too late. 😜 (re @Sannita: no, it was Nicolas who did) [12:38:52] Can we anticipate any functions that Dagbani Wikipedians will want to use? (And maybe make sure they're ready??) The announcement suggests: "This new capability will help volunteers perform tasks such as adding up numbers, converting distances, calculating time differences, or even transforming words into different grammatical cases." (re @Nicolas: Wikifunctions will [12:38:52] be deployed [12:38:52] on the Dagbani Wikipedia on April 15! [12:38:54] see https://lists.wikimedia.org/hyperkitty/list/wikimedia-l...) [12:41:11] @Sannita in case Denny is too busy, this FOTW would be timely for this week's news. (re @u99of9: I've added a FOTW submission for Z23254 and Z23269 a bit ahead of time aiming for the last newsletter before Easter. Feel free t...) [12:42:04] Yeah, I'd like to know that. (re @u99of9: Can we anticipate any functions that Dagbani Wikipedians will want to use? (And maybe make sure they're ready??) The announcemen...) [12:42:38] I'll suggest it for this week (re @u99of9: @Sannita in case Denny is too busy, you might be able to help: this FOTW would be timely for this week's news.) [12:48:28] Are there any Dagbani-speaking functioneers? 🤔 (re @u99of9: Can we anticipate any functions that Dagbani Wikipedians will want to use? (And maybe make sure they're ready??) The announcemen...) [12:50:25] it looks like Feeglgeef has researched the language more than I have: Z20618 Z20622 (re @Al: Are there any Dagbani-speaking functioneers? 🤔) [12:50:45] I have a very basic understanding (re @Al: Are there any Dagbani-speaking functioneers? 🤔) [12:51:51] maybe you can coordinate with @Sadike25 [12:52:15] good question, not sure how we could help best [12:52:16] I see these pages tho: https://www.wikifunctions.org/wiki/Category:Dagbani_Wikipedia_functions and https://www.wikifunctions.org/wiki/Wikifunctions:Catalogue/Natural_language_operations/Dagbani (re @u99of9: Can we anticipate any functions that Dagbani Wikipedians will want to use? (And maybe make sure they're ready??) The announcemen...) [12:55:03] Z20767 also (re @u99of9: it looks like Feeglgeef has researched the language more than I have: Z20618 Z20622) [12:57:40] @Dnshitobu too (re @Sannita: maybe you can coordinate with @Sadike25) [12:58:28] right, he's here too (re @Nicolas: @Dnshitobu too) [12:58:44] I forgot how many Dagbani users are here :D [12:58:50] Sure 👍 (re @Sannita: maybe you can coordinate with @Sadike25) [12:59:09] If we had a search like https://www.wikifunctions.org/w/index.php?limit=50&namespace=0&title=Special:WhatLinksHere/Z1015, we could see which functions a Dagbani speaker might discover… (re @Nicolas: good question, not sure how we could help best [12:59:10] I see these pages tho: https://www.wikifunctions.org/wiki/Category:Dagbani_Wikipe...) [13:00:50] These aren't particularly useful imo. Functions should save contributor effort. (re @wikilinksbot: Z20618 – dagbani possessive [13:00:51] Z20622 – negate dagbani verb) [13:01:11] Something like intro for year is much better for that [13:01:31] All of the items in the category should also be labeled in Dagbani iirc (re @Nicolas: good question, not sure how we could help best [13:01:33] I see these pages tho: https://www.wikifunctions.org/wiki/Category:Dagbani_Wikipe...) [13:01:52] The age function https://www.wikifunctions.org/wiki/Z20756 (re @u99of9: Can we anticipate any functions that Dagbani Wikipedians will want to use? (And maybe make sure they're ready??) The announcemen...) [13:02:41] That one isn't even usable since the AW team has refused to connect a renderer for dates (re @Sadike25: The age function https://www.wikifunctions.org/wiki/Z20756) [13:02:51] we're working on it [13:02:55] K [13:02:59] it might take a bit to make it work [13:03:11] sorry for suggesting a broken function :/ [13:03:13] 🤷‍♂️ I’m just not finding many Dagbani labels https://www.wikifunctions.org/w/index.php?ns0=1&ns4=1&search=%3A+Z8K2+Z1015&title=Special%3ASearch [13:14:24] I just took another look at Z20780. It looks decent, but I have a query and a suggestion. The query is that Z20770 seems to add a lot of extra words, and Google translate didn't tell me anything useful about them. The suggestion is that Z20788 would be better replaced IMO with an unambiguous format like 2025-05-10. Ne (re @Feeglgeef: That one isn't even usable since [13:14:24] the AW team [13:14:25] has refused to connect a renderer for dates) [13:18:46] parser* (re @Feeglgeef: That one isn't even usable since the AW team has refused to connect a renderer for dates) [13:19:33] Usually the reader/parser gets connected at the same time, so we're looking at Z20808. I've just added a couple of tests if anyone wants to work on broadening its coverage. (re @Feeglgeef: That one isn't even usable since the AW team has refused to connect a renderer for dates) [13:20:38] Z20770 is correct (re @u99of9: I just took another look at Z20780. It looks decent, but I have a query and a suggestion. The query is that Z20770 seems to add ...) [13:20:42] Oh yeah, the display is already connected. (re @Feeglgeef: parser*) [13:20:47] I'm pretty sure if that (re @Feeglgeef: Z20770 is correct) [13:22:19] Google translate doesn't even support Dagbani (re @u99of9: I just took another look at Z20780. It looks decent, but I have a query and a suggestion. The query is that Z20770 seems to add ...) [13:23:15] I believe this is the broadest search possible: https://www.wikifunctions.org/w/index.php?limit=100&ns0=1&ns4=1&offset=0&search=%3A+%22Z31K1+Z1015%22+OR+%22Z11K1+Z1015%22&title=Special:Search It looks for any labels or aliases for Z1015 in the main namespace. (re @Nicolas: good question, not sure how we could help best [13:23:16] I see these pages tho: https://www.wikifunctions.org/wiki/Category:Dagbani_Wikipe...) [13:28:23] There are lots more labels on Wikidata - perhaps we can think of some potentially useful functions to get info from there in Dagbani? [13:29:35] in that case, we should make sure the parser can accept those long strings and turn them back into dates (re @Feeglgeef: Z20770 is correct) [13:33:01] I have purely philosophical objections 🤔 (re @u99of9: There are lots more labels on Wikidata - perhaps we can think of some potentially useful functions to get info from there in Dag...) [13:35:01] Really? I'd love to hear them. I actually prefer the facts to come from the data, and the functionality to come from the functions. (re @Al: I have purely philosophical objections 🤔) [13:38:42] I just think Wikifunctions labels should be provided by people who know the language. (re @u99of9: Really? I'd love to hear them. I actually prefer the facts to come from the data, and the functionality to come from the functio...) [13:48:39] Oh, totally, that's fine. (re @Al: I just think Wikifunctions labels should be provided by people who know the language.) [13:58:45] Okay… but I don’t know anything about Dagbani syntax (yet). This is that missing step where we might need to go from the label to the lexeme to get the right form. (re @u99of9: Oh, totally, that's fine. I was meaning calling functions to get Wikidata statements in Dagbani, rather than copying labels.) [15:46:24] The format used on dagwiki today is “*15 Silimin gɔli April 2025*”, whereas the Python gives “silimin gɔli 15 April 2025 yuuni”. A date from Dagbani content is “29 September yuuni 1977”. My best guess is that “Silimin gɔli” means European month and “yuuni” means year, so 15 [European month] April [year] 2025 would seem to make sense 🤷‍♂️ (re @Feegl [15:46:24] geef: I'm pretty sure of that) [19:21:32] "Silimin" is English (re @Al: The format used on dagwiki today is “15 Silimin gɔli April 2025”, whereas the Python gives “silimin gɔli 15 April 2025 yuuni”. A...) [19:31:43] I wouldn’t know: [19:31:43] “Siliminsili n.pr. English, "European language". See: Silimina,. [19:31:45] Silimina, Pl: Siliminsi. Pl: Siliminsi. n.pr. 1 • a European. [19:31:46] 2 • white-man, non-African (may include Middle-Easterners, East and South Asians &c.) Syn: gbampielli, [19:31:48] Silimiga. Jaamani gbampiela (Siliminsi) ni daa yina ti yaanim saha. The German whitemen came in our ancestors' time. (DB-Bushfire.054) Syn: Nasara,, batule. Dagbamba (Gbansabila) yoli yeligu ka Silimiiga (Anashaara) deei zaa zani. The Dagomba (Africans) were late in getting to know things, so the Europeans go all the good positions. (Prov. 810) Sim: faada; Cpart: [19:31:48] gbansabinli. [19:31:49] 3 • a person of the Fulani tribe. Sim: Fulani. From: (Moore). Note: comes from Moore Filmi, Silmi "Peoples of Fular origin assimilated to Mossi" - ultimately from "Fulani" (pale-skinned people) ! See: Fulani; siliminaz.” [19:31:51] http://docs.wixstatic.com/ugd/ae2814_96c112f4dbdc4167bd0bc4bdea3826e4.pdf (re @Feeglgeef: "Silimin" is English) [20:13:22] You are correct ✅ (re @Al: The format used on dagwiki today is “15 Silimin gɔli April 2025”, whereas the Python gives “silimin gɔli 15 April 2025 yuuni”. A...) [20:17:59] That's a true analogy (re @Al: The format used on dagwiki today is “15 Silimin gɔli April 2025”, whereas the Python gives “silimin gɔli 15 April 2025 yuuni”. A...) [20:24:06] 🎉🎉🎉 Wikifunctions is now live on Wikipedia in Dagbani! 🎉🎉🎉 [20:24:07] See the message at the Dagbani WP's Community Portal (https://dag.wikipedia.org/wiki/Wikipedia:Mabiligu_dund%C9%94%C5%8B#c-Sannita_(WMF)-20250415202200-Sannita_(WMF)-20250414152700) [20:24:11] (re @Sannita: 🎉🎉🎉 Wikifunctions is now live on Wikipedia in Dagbani! 🎉🎉🎉 [20:24:12] See the message at the Dagbani WP's Community Portal) [20:24:21] @Sadike25 @Dnshitobu FYI [20:27:38] if you're interested in testing out how to include functions in articles, there's test.wikipedia.org for doing the testing, please do not use Wikipedia in Dagbani as a testbed [21:06:06] Well what do you know. [21:06:10] It actually works. [21:06:13] https://dag.wikipedia.org/wiki/%C5%8Aun_su:Amire80/Z [21:06:42] Use testwiki for testing please (re @amire80: https://dag.wikipedia.org/wiki/%C5%8Aun_su:Amire80/Z) [21:07:30] Oh, I made that edit before I saw what @Sannita wrote. [21:07:47] I can delete it if it causes trouble. [23:19:53] Thank you. Is that format good, or just a compromise? I mostly see years preceded by “yuuni” in articles. (re @Sadike25: You are correct ✅) [23:54:53] I copied the parent page to https://test.wikipedia.org/wiki/Wikifunctions/GrounderUK and fixed one of the function calls. These look like cached errors but they’re not showing up on Wikifunctions.org.