[03:17:29] Since T397494 is now complete we have Z6896 which takes a ZID string as input. Is there a way to make a composition for Z26879 which calls this? I briefly played with a few ZID functions, but couldn't get there. [07:04:40] Kind of… but the result is a list of Z6091s, not a list with the enumeration’s type. I’m not sure that matters, but the specific tests fail. (re @u99of9: Since T397494 is now complete we have Z6896 which takes a ZID string as input. Is there a way to make a composition for Z26879 w...) [07:09:15] True. We'll have to map them into the type with Z26779. I can't see your composition. How did you get from the type to the ZID of the type? (re @Al: Kind of… but the result is a list of Z6091s, not a list with the enumeration’s type. I’m not sure that matters, but the specific...) [07:23:30] I’m using Z22764 (converted to uppercase). I could try wrapping it with an apply2 🤔 [07:26:46] Now I've done in Z28505 (re @u99of9: Since T397494 is now complete we have Z6896 which takes a ZID string as input. Is there a way to make a composition for Z26879 w...) [07:28:30] 👍 that’s the one ✅ (re @dvd_ccc27919: Now I've done in Z28505) [07:31:11] The unwrapped version worked for Chemical element. (re @u99of9: True. We'll have to map them into the type with Z26779. I can't see your composition. How did you get from the type to the ZID o...) [08:49:14] Shall we call that a raw list? I imagine some compositions may operate faster on that. (re @Al: The unwrapped version worked for Chemical element.) [09:05:03] I’m sure you’re right. I’m not sure about the terminology but it’s always a list of Wikidata references, isn’t it? A “wdRef-list”, perhaps? 🤷‍♂️ (re @u99of9: Shall we call that a raw list? I imagine some compositions may operate faster on that.) [10:20:08] I love the enums 😍 (re @u99of9: Since T397494 is now complete we have Z6896 which takes a ZID string as input. Is there a way to make a composition for Z26879 w...) [12:12:48] I haven't put Z28509 through its paces, but my first shot at a composition seems to be working okay already, and would certainly be fine for metres to kilometres. (re @Al: Wonderful! I’ll gladly leave the “fancy precision stuff” to you, but I’m happy to help in other ways.) [13:34:37] I’ve given Z28514 quite a hammering and it seems to hold up. I’ll give it some tests in due course… [16:02:17] This test is using live data from Wikidata. Z28522 [16:02:18] I find that suboptimal. Has anyone else brought that up? [16:07:24] I prefer to use live data, not least because emulating a Wikidata statement is far from straightforward. If the data changes, the test will fail, of course. (re @Npriskorn: This test is using live data from Wikidata. Z28522 [16:07:25] I find that suboptimal. Has anyone else brought that up?) [16:20:56] ok so given all things its the least terrible we have come up with? (re @Al: I prefer to use live data, not least because emulating a Wikidata statement is far from straightforward. If the data changes, th...) [16:27:54] I don’t recall hearing any better idea so far. You can try fixing Z23723, if you like. There’s an issue with replicating the omitted keys we have in authentic statements fetched from Wikidata. (re @Npriskorn: ok so given all things its the least terrible we have come up with?)