[00:41:34] I tried to get around this, at the same time preserving the type of the original list with Z17708, but clearly do not understand how typed lists work in code... :-( (re @Al: My guess is Z12810, which fails for a natural number list, with a bigint error.) [01:32:37] So "let result = Z12668K1;" just points to the same location, it doesn't make a copy. Meanwhile I haven't found a good way to make a copy that works for this. (re @Toby: I tried to get around this, at the same time preserving the type of the original list with Z17708, but clearly do not understand...) [04:02:40] Thanks. I've resolved this issue too, so the outcome is now symmetric and obeys this convention. (re @Al: It looks okay for the given tests, but I think gcd(42, 0) = gcd(0, 42) = 42 (by convention, not according to Euclid or his algor...) [04:20:55] COBOL? (re @Hogü-456: How many characters are allowed in a row of a code implementation of a function in Wikifunctions. I try to parse the Wikifunctio...) [04:21:13] Seems this is an ancient programming language ! [05:35:51] 6 million views! : https://tools-static.wmflabs.org/bridgebot/84831b65/file_62554.jpg [05:44:03] https://tools-static.wmflabs.org/bridgebot/77997b03/file_62555_oga.ogg [05:44:16] https://tools-static.wmflabs.org/bridgebot/94b71f93/file_62556_oga.ogg [05:45:11] This is so cool. Let the best/fastest/least failing implementations see the widest adoption 🤩 (re @vrandecic: That's the neat thing: you can go and create another implementation, without having to even remove or delete the existing one!) [06:34:26] crazy spiders :) : https://tools-static.wmflabs.org/bridgebot/939ef40d/file_62557.jpg [09:13:01] Strictly speaking, they don’t (because the arrays are not typed). See the errors for the (unconnected) JavaScript and Python implementations for Z801 with test Z13870. These suggest quite strongly that there is an error converting from an array of integers to a Typed list (when the type is not Z1). (re @Toby: I tried to get around this, at the same time preserving the [09:13:01] type of t [09:13:01] he original list with Z17708, but clearly do not understand...) [10:04:47] *T370028* (re @Toby: I tried to get around this, at the same time preserving the type of the original list with Z17708, but clearly do not understand...) [13:18:10] Thanks for the report. I have trouble knowing when the brick wall I'm hitting is due to my ignorance or a bug or both. (re @Al: T370028) [13:30:50] So if Python returns an array containing positive integers, how can WF know what type to make the objects within the list? They could be integers, natural numbers, months, days, igbo months, ... It seems like we need a way to specify the list element types? (re @Al: Strictly speaking, they don’t (because the arrays are not typed). See the errors for the (unconnected) [13:30:50] JavaScript [13:30:51] and Python imp...) [13:30:52] Or maybe we need to break up functions like Z12668 into one for each type, e.g. "reverse list of natural numbers". Would that work? (re @Al: Strictly speaking, they don’t (because the arrays are not typed). See the errors for the (unconnected) JavaScript and Python imp...) [13:40:54] Perhaps a "reverse list (objects)" would be better done as a built-in function only? [13:43:06] That’s where I was heading with my comment on https://www.wikifunctions.org/wiki/Wikifunctions:Project_chat#Typed_lists_now_open_beyond_Booleans_and_Strings. I think there must already be inference based on the input but I can’t think of an example that demonstrates this. (re @Toby: So if Python returns an array containing positive integers, how can WF know what type to [13:43:06] make [13:43:07] the objects within the list? They ...) [13:48:56] Lists of elements of mixed types is next-level! (re @Al: That’s where I was heading with my comment on https://www.wikifunctions.org/wiki/Wikifunctions:Project_chat#Typed_lists_now_open...) [14:00:22] True, but my starting point was lists of elements of one (unspecified) type. Z13755 returns a Boolean list when given an untyped list containing only Boolean elements (see T359500#9610983). (re @Toby: Lists of elements of mixed types is next-level!) [16:36:32] I want to watch all functions, impl and tests I create and edit. I set this in the preferences but none of the pages I edit gets added. Is this a known issue? : https://tools-static.wmflabs.org/bridgebot/1fcb00ab/file_62576.jpg [16:37:30] huh, that's weird (re @dpriskorn: I want to watch all functions, impl and tests I create and edit. I set this in the preferences but none of the pages I edit gets...) [16:37:37] does anybody else have the same problem? [16:38:01] maybe it's worth it to create a ticket on Phabricator and link it back to us, so that we might look into it [16:42:41] Seem to be the same for me (re @Sannita: does anybody else have the same problem?) [17:20:54] Could a functioneer please disconnect Z17704 so I can edit it and make the last test pass on the function? [17:21:41] (I applied for functioneer right yesterday) [17:23:28] Done ✅ (re @dpriskorn: Could a functioneer please disconnect Z17704 so I can edit it and make the last test pass on the function?) [18:37:29] Is there a way to define a function where the output is a Wikidata item. [18:37:40] Just a simple question. [18:39:07] Not for now (re @Csisc1994: Is there a way to define a function where the output is a Wikidata item.) [18:39:34] Would be available later if we have better Wikidata integration [20:20:40] I think that we should create a Python Package that loads and uses Wikifunctions implementations. [20:21:14] I think that we should also upload all medical functions to Wikifunctions. [20:21:53] This is not very difficult [20:21:57] Thank you