[01:21:00] Why can't we just deploy wf on enwiki? [03:52:38] Please no! We have way too much to learn before then. (re @Feeglgeef: Why can't we just deploy wf on enwiki?) [04:20:41] Anyone want to help debug this for me? It's a super simple composition, and each of the steps seems to work fine independently: Z24043 [04:35:03] Something for you, if we enter a the qid as a 'literal string' in the composition, then it works. So something wrong in the weekday to qid function output. : https://tools-static.wmflabs.org/bridgebot/d98df4ff/file_69981.jpg [04:35:19] Btw I connected this for temporary testing, hope you don't mind! (re @u99of9: Anyone want to help debug this for me? It's a super simple composition, and each of the steps seems to work fine independently: ...) [07:13:49] @u99of9 It's working now. I removed the Z6091 wrapper for QID param of the first function as it was given by Z23419 [07:15:08] but for some reason it's not saving. [07:15:13] 🤔 I gave the inner function a new implementation (composition) and disconnected the JavaScript. Seems to be working but I haven’t reset the cache. (re @u99of9: Anyone want to help debug this for me? It's a super simple composition, and each of the steps seems to work fine independently: ...) [07:26:04] Works in edit menu, but test fails after "save" (which does not happen). Is it also a cache issue? : https://tools-static.wmflabs.org/bridgebot/3d989db8/file_69982.jpg [07:44:40] I tried reconnecting the JavaScript but the outer function’s tests failed, so I disconnected it again and reset the cache for Z24041. (re @u99of9: Anyone want to help debug this for me? It's a super simple composition, and each of the steps seems to work fine independently: ...) [08:04:07] Thank you so much (re @u99of9: Welcome Eunice and Faatimah) [08:06:25] Is there a way to put a test on the inner function to show that it causes overall failure? Otherwise someone forgetful like me might reenable it. (re @Al: I tried reconnecting the JavaScript but the outer function’s tests failed, so I disconnected it again and reset the cache for Z2...) [08:08:14] I’ll give it a go 🤔 (re @u99of9: Is there a way to put a test on the inner function to show that it causes overall failure? Otherwise someone forgetful like me m...) [08:54:04] I tried roundtripping with the inverse function, but it passed 🤷‍♂️ Z24048. I could try creating a “compare after apply2” function or, failing that (the simpler option) “label of item reference in English” 🤔 (re @u99of9: Is there a way to put a test on the inner function to show that it causes overall failure? Otherwise someone forgetful like me m...) [10:48:32] That would be a no… I created Z24053 with new function Z24049 but Z23421 keeps passing (and there’s no apparent reason why it shouldn’t, of course) 🤷‍♂️ (re @u99of9: Is there a way to put a test on the inner function to show that it causes overall failure? Otherwise someone forgetful like me m...) [13:28:46] And finally… I’ve added Z24054 to Z24041 and re-connected Z23421. All tests now fail for Z24043, which remains disconnected: “Call tuples failed in returnOnFirstError. Error: TypeError: s.match is not a function.” (Phabricator ticket to be filed in due course.) (re @u99of9: Anyone want to help debug this for me? It's a super simple composition, and each of the steps seems [13:28:47] [13:28:48] to work fine independently: ...) [13:35:34] Well done, that is very comprehensive. Please subscribe me to the ticket. [13:39:21] 👍 I’m sure I wanted Z24049 before, but I can’t remember where 😁 (re @u99of9: Well done, that is very comprehensive. Please subscribe me to the ticket.) [13:46:43] It will happen, but later, much later, also because en.wp has less reason to adopt Wikifunctions than a small community that struggles with adopting en.wp articles, functions are more useful on smaller projects where people do not have the possibilities of bigger projects (re @Feeglgeef: Why can't we just deploy wf on enwiki?) [21:01:49] Update: I think I worked out what the problem is/was. I have created an alternative JavaScript implementation that returns Z6091K1 as a simple string rather than a String object Z24071. If the String object in a constructed reference object is unwrapped, it can be interpreted as a Typed Map and cause an Argument type mismatch. I’m holding off on filing the Phabricator [21:01:49] ticket fo [21:01:49] r now. (re @Al: And finally… I’ve added Z24054 to Z24041 and re-connected Z23421. All tests now fail for Z24043, which remains disconnected: “Ca...)