[00:10:42] I most humbly beg forgiveness from the type-safety gods Z23679 (re @u99of9: Now I've got the minimal viable Z23630 going you can tell me if it's abusing every principle of strong typing.) [00:12:11] (Actually it’s a Natural number) (re @Al: I most humbly beg forgiveness from the type-safety gods Z23679) [00:16:01] Should unproductive half-complete implementations with little likely prospect of rehabilitation be kept or deleted? Z23633 [00:17:23] Yes. At the discretion of the author and administrator 😎 (re @u99of9: Should unproductive half-complete implementations with little likely prospect of rehabilitation be kept or deleted? Z23633) [00:18:47] Deletion should be done as long as there is no objection (re @u99of9: Should unproductive half-complete implementations with little likely prospect of rehabilitation be kept or deleted? Z23633) [00:19:09] If there is any objection it shouldn't be deleted unless under exceptional circumstances [03:51:24] Error persisted, so I purged. Can we have a serious think about purging cached errors in display and read functions, please? (re @Al: Still getting display error on Integer –1… shall I purge?) [05:20:24] Test Z23690 is missing a subject. I have been trying to set it to Q7055322, but no matter where I put it or how I type it in, it doesn't persist long enough to save. Please don't add it in manual mode, I want to investigate if there is an issue with the normal interface. [06:13:21] Not really? [06:13:23] `Z23691` [06:13:24] Feel free to delete if this answers your question. (re @u99of9: Test Z23690 is missing a subject. I have been trying to set it to Q7055322, but no matter where I put it or how I type it in, it...) [12:54:25] Can you change the original one without entering manual edit mode? (If you can, maybe don't press publish). Here is me trying again (from a different computer): : https://tools-static.wmflabs.org/bridgebot/9ded37b6/file_69708.mp4 [13:09:01] It has to be changed to a Wikidata item reference. It’s currently set as a function call to Fetch… In a new test it is not so set (as in Try this…); you have to choose the reference type, not Wikidata item. (re @u99of9: Can you change the original one without entering manual edit mode? (If you can, maybe don't press publish). Here is me trying ag...) [13:10:27] MediaWiki internal error. [13:10:27] Original exception: [315c3b82-69a0-40d4-a7bb-bc0a2cf97ad7] 2025-04-04 13:09:48: Fatal exception of type "Wikimedia\Rdbms\DBUnexpectedError" [13:10:29] Exception caught inside exception handler. [13:10:30] Set $wgShowExceptionDetails = true; at the bottom of LocalSettings.php to show detailed debugging information. [13:10:39] at https://www.wikifunctions.org/view/en/Z23690 (re @u99of9: MediaWiki internal error. [13:10:39] Original exception: [315c3b82-69a0-40d4-a7bb-bc0a2cf97ad7] 2025-04-04 13:09:48: Fatal exception of ty...) [13:12:05] Is that on Publish? (re @u99of9: MediaWiki internal error. [13:12:06] Original exception: [315c3b82-69a0-40d4-a7bb-bc0a2cf97ad7] 2025-04-04 13:09:48: Fatal exception of ty...) [13:12:38] No, it was just transient as I went back to the page from Telegram to follow your instructions. But then I refreshed and it was fine. (re @Al: Is that on Publish?) [13:13:48] Oh, but it's now reproducible. I click the top link in this item and just press open. (re @wikilinksbot: Z23690 – normal rank statement is not preferred [13:13:48] Q7055322 – North Epping) [13:14:58] Not anymore, but it happened three times. (re @u99of9: Oh, but it's now reproducible. I click the top link in this item and just press open.) [13:15:29] It seems a wider problem, not just 'functions (re @u99of9: MediaWiki internal error. [13:15:29] Original exception: [315c3b82-69a0-40d4-a7bb-bc0a2cf97ad7] 2025-04-04 13:09:48: Fatal exception of ty...) [13:15:48] Yes, I got it once, after you said 🤷‍♂️ (re @u99of9: Not anymore, but it happened three times.) [13:18:35] Oh right, I chose subject Type=Wikidata item, instead of item ref. That inserts the fetch function. Thanks. I bet I fall into this trap again! (re @Al: It has to be changed to a Wikidata item reference. It’s currently set as a function call to Fetch… In a new test it is not so se...) [13:20:45] Yeah, the usability is not great, because there is no visual distinction unless you expand. So do that! 😎 (re @u99of9: Oh right, I chose subject Type=Wikidata item, instead of item ref. That inserts the fetch function. Thanks. I bet I'll fall into...) [13:21:29] But weirdly I made the same mistake in the value field, and met no obstacle (it's still there now after I fixed the subject) (re @u99of9: Oh right, I chose subject Type=Wikidata item, instead of item ref. That inserts the fetch function. Thanks. I bet I'll fall into...) [13:22:52] I expanded many times. The error was that I was aiming for item-prop-item. I'd forgotten that they all had to be IDs. (re @Al: Yeah, the usability is not great, because there is no visual distinction unless you expand. So do that! 😎) [13:24:27] https://tools-static.wmflabs.org/bridgebot/bb319060/file_69711.jpg [13:24:48] Maybe it’s not supposed to matter 🤷‍♂️ I only know what works 😎 (re @u99of9: But weirdly I made the same mistake in the value field, and met no obstacle (it's still there now after I fixed the subject)) [13:26:37] One way to hint it better would be to change the key labels in Z6003. E.g. "subject QID/LID/LFID/LSID" (re @Al: Yeah, the usability is not great, because there is no visual distinction unless you expand. So do that! 😎) [13:33:40] One for David… presumably the value types have to be Z1 to allow for the different reference types, but, yeah, that’s a bit unfortunate. (re @u99of9: One way to hint it better would be to change the key labels in Z6003. E.g. "subject QID/LID/LFID/LSID") [13:38:47] Also, the IDs are not searchable, so it's actually easier to input an item. (re @u99of9: I expanded many times. The error was that I was aiming for item-prop-item. I'd forgotten that they all had to be IDs.) [13:39:48] They’re searchable if you collapse the selector, aren’t they? (re @u99of9: Also, the IDs are not searchable, so it's actually easier to input an item.) [13:42:40] Oh yes, nevermind, that is good behaviour. (re @Al: They’re searchable if you collapse the selector, aren’t they?) [21:46:42] Yes, the type for subject and for value has to be Z1, to allow for the different types. One day we'll be able to do better, when we have union types, which is a desire that comes up pretty frequently. I like the hint suggestion, but wondering if one of these might be a bit better: "subject (Wikidata entity reference)" or "subject (Z6091/2/4/5/6)". Not thrilled with any [21:46:42] of t [21:46:42] hem, but I think I like "subject (Wikidata entity reference)" best. (re @Al: One for David… presumably the value types have to be Z1 to allow for the different reference types, but, yeah, that’s a bit unfo...) [22:33:26] That works for me. (re @David: Yes, the type for subject and for value has to be Z1, to allow for the different types. One day we'll be able to do better, whe...)