[10:09:35] *Abstract Wikipedia Newsletter 247* [10:09:35] * References from Wikidata now available [10:09:36] * Recording of Toby Hudson’s talk at WikiCon Australia [10:09:38] * Abstract data dashboard [10:09:39] * Admins on Abstract Wikipedia? [10:09:41] * Recent Changes in the software [10:09:42] * Volunteers’ Corner on May 11 [10:09:44] * Fresh Functions weekly: 39 new Functions [10:09:45] https://www.wikifunctions.org/wiki/Wikifunctions:Status_updates/2026-05-08 [12:49:03] https://tools-static.wmflabs.org/bridgebot/be5fdc3c/file_80091.jpg [12:49:20] When I search for the string "עברית" (it means "Hebrew" in Hebrew), I get two results: [12:49:31] Hebrew and Biblical Hebrew [12:51:49] Where does the Hebrew label of Biblical Hebrew come from? [12:52:16] The label at https://www.wikifunctions.org/view/en/Z1667 is not the same. [12:55:11] Isn't it from this fragment? : https://tools-static.wmflabs.org/bridgebot/7e4e777d/file_80092.jpg [12:55:29] The first word is the same, the second word is different. (re @u99of9: Isn't it from this fragment?) [12:57:44] I think the search is just looking to match at least the fragment typed. So if you just type "biblical" then you still see a result, because it is part of the full string "biblical Hebrew" [12:58:09] I understand that, but I don't understand where does that second word come from. [12:59:34] Ah sorry, yes I can see the issue. [13:01:27] Oh, I think I found it. [13:01:46] It's an alias in Hebrew, but it's hidden by default. [13:02:02] עברית תנ"כית [13:02:03] Is a Hebrew alias for Biblical Hebrew. [13:02:32] But this still begs the question: Why is this alias shown in the autocomplete box and not the label, which begins with the same word? [13:04:08] The label should be prioritized over the alias if it begins with the same characters, isn't it? [13:05:10] That sounds sensible (re @amire80: The label should be prioritized over the alias if it begins with the same characters, isn't it?) [18:50:12] I was wandering: in the future, Abstract Wikipedia articles will probably need not to generate single isolated fragments, but intere paragraphs if not entire articles in a single function call (since the rendering of abstract text usually depends on the text that has already been generated, for example for handling pronouns or for putting wikilinks only the [18:50:12] first time a concept i [18:50:12] s mentioned). How could we make it possible to do it? Now it would be impossible, due to the really high execution times that such a function call would require [19:14:30] I mean generating all-in-one works in some cases right now, I think improvements in speed will mean it's ok (re @dvd_ccc27919: I was wandering: in the future, Abstract Wikipedia articles will probably need not to generate single isolated fragments, but in...) [20:11:19] Do you have any example? For now, every all-in-one functions I've seen in Abstract Wikipedia don't work for Italian (where it is necessary to fetch lexemes for each item, due to grammatical genders) (re @Feeglgeef: I mean generating all-in-one works in some cases right now, I think improvements in speed will mean it's ok)