[06:06:07] Any ideas for what should be Z40000? I was thinking perhaps auto-infobox which would infer which particular infobox to use based on the P31 of the QID. Other ideas welcome. [06:07:28] (it's in 8 ZIDs time, so speak soon!) [06:34:02] That's a good idea (re @u99of9: Any ideas for what should be Z40000? I was thinking perhaps auto-infobox which would infer which (if any) particular infobox to ...) [06:35:01] It will simplify the work of editors [06:35:38] Great idea! (re @u99of9: Any ideas for what should be Z40000? I was thinking perhaps auto-infobox which would infer which (if any) particular infobox to ...) [06:37:04] I'm afraid we will almost certainly run into trouble fighting the way of reserving good numbers. Idea: Reserve the whole Z4XXXX block (or other easily memorable block) for very important functions; tell the system to skip them when automatically assigning function IDs, and allow creation within these reserved blocks by sysops. [06:38:49] All numbers in a block are not easily memorable though. (If so, every number in all blocks would be.) (re @nya_1F616EMO: I'm afraid we will almost certainly run into trouble fighting the war of reserving good numbers. Idea: Reserve the whole Z4XXXX ...) [06:40:19] Okay, seeing support for this, I'll contrive to make it happen in a few minutes. (re @u99of9: Any ideas for what should be Z40000? I was thinking perhaps auto-infobox which would infer which (if any) particular infobox to ...) [06:40:46] Btw, support on making that number the infobox selection function. [06:48:02] Boom Z40000! Now we've just got to make it work šŸ˜„ [07:01:34] Awesome. Thanks (re @u99of9: Boom Z40000! Now we've just got to make it work šŸ˜„) [07:21:35] @jsamwrites (and anyone else who wants), are you interested in trying to expand this list to include all the infobox functions you've made? With it, I think I can then construct the implementation. Z40008 [07:27:44] Perhaps all Z*000? (re @nya_1F616EMO: I'm afraid we will almost certainly run into trouble fighting the war of reserving good numbers. Idea: Reserve the whole Z4XXXX ...) [07:31:18] Sure (re @u99of9: @jsamwrites (and anyone else who wants), are you interested in trying to expand this list to include all the infobox functions y...) [07:55:26] Done. (re @u99of9: @jsamwrites (and anyone else who wants), are you interested in trying to expand this list to include all the infobox functions y...) [07:57:19] I'm on a train but will try to get the implementation done tonight. [08:48:51] That would make more sense [08:48:52] An alternative is to make redirects in human language (e.g. [[INFOBOX]] would then point to [[Z40000]]), so that we can search for functions more easily (re @u99of9: Perhaps all Z*000?) [08:56:16] Isn't this what [[Wikifunctions:Catalogue]] is for? (re @nya_1F616EMO: That would make more sense [08:56:16] An alternative is to make redirects in human language (e.g. [[INFOBOX]] would then point to [[Z40000...) [09:36:16] [[:abstract:Abstract Wikipedia talk:About#How does Abstract Wikipedia work?]] (re @Al: Yes, I think thatā€™s the broad shape of the plan. I should try to articulate something on-wiki, but I wouldnā€™t want to appear to ...) [09:49:56] ā€œinfoboxā€ is an English alias for Z40000, so searching for ā€œinfoboxā€ surfaces Z40000 (currently after typing ā€œinfobā€). (re @nya_1F616EMO: That would make more sense [09:49:57] An alternative is to make redirects in human language (e.g. [[INFOBOX]] would then point to [[Z40000...) [10:17:51] it even works on Q40000 https://abstract.wikipedia.org/wiki/Q40000 : https://tools-static.wmflabs.org/bridgebot/a3cf939b/file_83335.jpg [10:44:35] In the Wikipedia app, infoboxes are collapsed by default. Is that something we can support for AW? I guess it must be a CSS thing? šŸ¤·ā€ā™‚ļø : https://tools-static.wmflabs.org/bridgebot/36b72107/file_83338.jpg [10:45:40] Don't you think the Wikipedia app devs will fix that themselves? (re @Al: In the Wikipedia app, infoboxes are collapsed by default. Is that something we can support for AW? I guess it must be a CSS thin...) [10:46:01] I mean, we didn't need to do anything on svwiki for it to happen... [10:47:22] I was thinking of ordinary mobile users of AW (which the app does not support). (re @Jan_ainali: Don't you think the Wikipedia app devs will fix that themselves?) [10:48:17] You mean users on abstract.wikipedia.org? [10:49:50] Yes. Presumably the client wikis will just work (or fail completely). (re @Jan_ainali: You mean users on abstract.wikipedia.org? (And not the client wikis?)) [10:50:49] Then I think we should _not_ collapse it. Being on AW itself is kind of being in the editor on the other projects. It's not for consuming content, it's for editing it. [10:53:24] Fair. But the text wraparound is aesthetically evil. (re @Jan_ainali: Then I think we should not collapse it. Being on AW itself is kind of being in the editor on the other projects. It's not for co...) [11:09:25] https://tools-static.wmflabs.org/bridgebot/fa40fc68/file_83339.jpg [11:36:33] If anyone is up for it, a code implementation would be very useful for Z40010 [11:37:05] When I click the refresh icon next to a fragment that renders as "Missing", nothing happens [11:38:02] which article are you on? some of them take a long time to fully recalculate. (re @Earldridge Jazzed: When I click the refresh icon next to a fragment that renders as "Missing", nothing happens) [11:38:39] Q928 [11:39:06] Occurs while I was editing the article, but the fragments do render properly once I publish my edit [11:40:34] Not sure then, sorry. [11:41:33] I was splitting an instance of Z32163 consisting of 3 fragments (prone to rate limiting errors) into 5 fragments (including 2 sentence separators). [11:42:55] Yes, at the moment it is better off split, for that reason. [11:44:43] When we get to it, the Simple English might be ā€œThe person it gets its name from is Kim Chaek.ā€ (Thatā€™s a very inclusive ā€œweā€, but Simple English is not my mother tongue!) I guess that should be the ordinary ā€œenā€ rendition too, although I think this works better, as a target: [11:44:43] ā€œThe person it gets its name from is North Korean general and politician (1903ļ¼1951).ā€ [11:44:45] (with the eponym being a link, of course). [11:44:46] Iā€™ve lifted the short description from Wikidata, which is easy enough, but we ought to construct them in AW. (re @Al: ) [11:45:03] A few of those functions could now be switched to their HTML versions, which would simplify the page. (re @Earldridge Jazzed: I was splitting an instance of Z32163 consisting of 3 fragments (prone to rate limiting errors) into 5 fragments (including 2 se...) [11:48:04] For that particular property, there is a competing option to use Z38443.It would be easier to provide a simple-specific structure for that one. (re @Al: When we get to it, the Simple English might be ā€œThe person it gets its name from is Kim Chaek.ā€ (Thatā€™s a very inclusive ā€œweā€, b...) [11:49:38] Probablyā€¦ yes. Thanks. (re @u99of9: For that particular property, there is a competing option to use Z38443.It would be easier to provide a simple-specific structur...) [11:49:53] Personally I'd cut out "the person". So just "It gets its name from [[Kim Chaek]]." And if that simple, it wouldn't need a link to eponym either. (re @Al: When we get to it, the Simple English might be ā€œThe person it gets its name from is Kim Chaek.ā€ (Thatā€™s a very inclusive ā€œweā€, b...) [11:52:56] I guess this would be a good target for HTMLification soon. A link and reference are really helpful for this kind of statement. (re @u99of9: For that particular property, there is a competing option to use Z38443.It would be easier to provide a simple-specific structur...) [11:53:50] Not a link to ā€œeponymā€, a link to the eponym, Kim Chaek. Iā€™m English, so Iā€™m happy with ā€œnamed afterā€, but Americans favo(u)r ā€œnamed forā€. ā€œIt gets its name from [[link]]ā€? (re @u99of9: Personally I'd cut out "the person". So just "It gets its name from [[Kim Chaek]]." or "It is named after Kim Chaek." And if tha...) [11:54:16] Ugh, it doesn't even have a good default function on Z38414 [11:57:49] Iā€™ll take a look at that too. (re @u99of9: Ugh, it doesn't even have a good default function on Z38414) [12:14:50] It looks a lot like Z14403. I think weā€™d be better off going into code with a (string,ZID) map. Z96 is reasonably imminent, isnā€™t it? (re @u99of9: If anyone is up for it, a code implementation would be very useful for Z40010) [12:17:12] Z14403 (re @Al: It looks a lot like Z14403. I think weā€™d be better off going into code with a (string,ZID) map. Z96 is reasonably imminent, isnā€™...) [12:18:37] ā€¦or (QID,ZID) šŸ¤” (re @Al: It looks a lot like Z14403. I think weā€™d be better off going into code with a (string,ZID) map. Z96 is reasonably imminent, isnā€™...) [12:31:18] There is some information that is hard to include in Q928. I could not find functions for "X has N Ys" ("X has Y" exists but only for English) or "Subject verb object" (though I vaguely remember such a function existing). Nor could insert an arbitrary string in place of Wikidata item references. [12:38:07] "X has N Ys" is a good one we should make. I've wanted it a few times as well. Subject verb object had theoretical problems because in some languages the subject and object are inverted. What are you hoping to do with arbitrary strings? How would they work abstractly in all languages? (re @Earldridge Jazzed: There is some information that is hard to include in Q928. I [12:38:08] could not f [12:38:09] ind functions for "X has N Ys" ("X has Y" exists but onl...) [12:39:04] Wow, I just looked at your article again. You've made a great improvement! Nice work. (re @Earldridge Jazzed: There is some information that is hard to include in Q928. I could not find functions for "X has N Ys" ("X has Y" exists but onl...) [12:42:00] Z40001 is using the wrong Sydney [12:44:27] Thanks, fixed. (re @nya_1F616EMO: Z40001 is using the wrong Sydney) [12:49:57] I vaguely remember there being a function that generated sentences of the form "s s". I wonder what happened to it? (re @u99of9: "X has N Ys" is a good one we should make. I've wanted it a few times as well. Subject verb object had theoretical problems beca...) [12:51:46] You're probably thinking of Z26955 (re @Earldridge Jazzed: I vaguely remember there being a function that generated sentences of the form "s s". I wonder what happ...) [13:13:17] That function's implementations only do SVO word order. Is there a replacement function that uses a config so that it works for any word order, such as SOV, VSO, or VOS? (re @u99of9: You're probably thinking of Z26955. If so, have a look at the talk page for background on why it was marked as DO NOT USE.) [13:14:00] Yes, I was thinking of a small nod towards [[:en:Referring expression generation]] with an explicit [current topic] QID. This is similar to link-removal (which I view as a special case of anaphora). It would generally be pre-emptive, and may lead to reduced (or eliminated) surface text in place of the fully realized noun phrase (ā€œitā€ or ā€œthe universityā€ instead of ā€œKim [13:14:00] [13:14:01] Chaek University of Technologyā€ā€¦ or consider ā€œLocated by [the] Taedong River, it was established in 1948ā€). (re @u99of9: "X has N Ys" is a good one we should make. I've wanted it a few times as well. Subject verb object had theoretical problems beca...) [13:16:31] Switching it around would just be up to each language, so that would be fine, the issue was roughly that while two languages sometimes agree about which word is the subject, in other sentences they disagree. (re @Earldridge Jazzed: That function's implementations only do SVO word order. Is there a replacement function that uses a config so that it works for ...) [13:23:51] Or the same language can arrange it in different ways. Itā€™s important that AW specifies the unambiguous relation, but a minimal discourse frame could help with fluency (and by ā€œminimalā€, I mean a single QID, initially). [13:24:56] what is a minimal discourse frame? [13:27:14] To allow for more complex sentences. For example, a theoretical article on Q16989129 can have this complex sentence: [13:27:15] "Darling is a compatibility layer for Linux that runs macOS software." [13:27:16] I could use Z32965, but "compatibility layer for Linux that runs macOS software" is not on Wikidata (it is too specific). It would be broken down like this: [13:27:18] * subordinate clause [13:27:19] ** "for" clause [13:27:21] *** indefinite noun phrase [13:27:22] **** compatibility layer [13:27:24] *** Linux [13:27:25] ** verb + object [13:27:27] *** run [13:27:28] *** noun phrase with adjective [13:27:30] **** macOS [13:27:31] **** software (re @u99of9: "X has N Ys" is a good one we should make. I've wanted it a few times as well. Subject verb object had theoretical problems beca...) [13:30:10] Then I would use the subordinate clause function as Z32965K2, instead of a Wikidata item reference. [13:30:45] A QID referencing the foregrounded topic. A less minimal discourse frame would allow related relations or multiple QIDs, but one step at a time! (re @u99of9: what is a minimal discourse frame?) [13:31:24] Oh, I see, you don't mean simple "strings". There are type proposals that seek to address this kind of generalisation: e.g. [[Wikifunctions:Type_proposals/Semantic_unit]] and [[Wikifunctions:Type_proposals/Words_with_many_forms]] but IMO they are still a fair way off easy use from AW. (re @Earldridge Jazzed: To allow for more complex sentences. For example, a [13:31:24] theoretical article [13:31:24] on Q16989129 can have this complex sentence: [13:31:25] "Darling is...) [13:36:22] Got it. I support starting to include a little context. @Jsamwrites uses a boolean "mentioned" instead. What would you pass into yours if there was no previously foregrounded item? (re @Al: Yes, I was thinking of a small nod towards [[:en:Referring expression generation]] with an explicit [current topic] QID. This is...) [13:38:54] Is there a way I can refresh the rendering of functions that say "Wikifunctions returned a failed response: Reached rate limit in orchestrator" in Abstract Wikipedia? Using Special:Purge did not work. [13:42:35] You're right, I am using the concept of entity repeated. Examples: Z38443 (re @u99of9: Got it. I support starting to include a little context. @Jsamwrites uses a boolean "mentioned" instead. What would you pass into...) [13:43:05] Z38448 [13:44:31] Not easily. There are phabricator tasks asking for this. For now you just need to wait or switch languages. (re @Earldridge Jazzed: Is there a way I can refresh the rendering of functions that say "Wikifunctions returned a failed response: Reached rate limit i...) [13:45:36] One trick you can use once per day is to switch from the "this article's subject" argument to a literal QID for that page. (re @Earldridge Jazzed: Is there a way I can refresh the rendering of functions that say "Wikifunctions returned a failed response: Reached rate limit i...) [14:21:14] Iā€™m not sure that Iā€™d pass anything. Iā€™m trying to avoid wrapping the whole function, but if we assume weā€™d have to, the way weā€™d avoid stating the current topic is by not having the wrapper that supplies it. Alternatively, we assume that the discourse frame is a list of objects and only initially support a minimal list containing 0 or 1 Z6091 (any other objects being i [14:21:14] [14:21:15] gnored, rather than an error). I lean towards the latter approach, in the hope that it would also support the envisaged new types. (re @u99of9: Got it. I support starting to include a little context. @Jsamwrites uses a boolean "mentioned" instead. What would you pass into...) [18:18:57] I'm incredibly excited and grateful for the opportunity to invest and receive a total profit of $43,000. This experience has been truly rewarding, and I'm happy with the outcome. If you're interested in learning about forex trading, I recommend you manager > LISA_FOREX (https://t.me/m/ov63tW_SYjA0)