[02:43:03] Is the failure of Z26568 another V2 hangover? [06:40:36] Anyone want to try to crack a code implementation of Z35809? It would have a shot at making Z35808 the most efficient implementation of the very important Z24766. [06:44:16] Is it even possible? There is the issue about references (re @u99of9: Anyone want to try to crack a code implementation of Z35809? It would have a shot at making Z35808 the most efficient implementa...) [07:08:28] Yes. Should we be requesting a Z46 for Z11 or for Z60 or something? If the mapping was done as part of a code conversion, we could rely on it? Would this work? (re @dvd_ccc27919: Is it even possible? There is the issue about references) [07:54:46] This is useful. I was thinking of as standardizing templates based on different topics (re @u99of9: @Dnshitobu I'm just listening to the Volunteer's Corner, and I heard your question about copying the structure of Australia to G...) [07:58:02] Regarding standard templates, I agree with Denny that countries are too different to make much use of that. There will be some topics where it makes a lot of sense, but I think we are not ready for them either, because we don't yet have a big enough vocabulary of sentences to express what we need for any topic. (re @Dnshitobu: This is useful. I was thinking of as [07:58:02] standardizing te [07:58:04] mplates based on different topics) [08:03:18] Changed to use Z14392 ✅ (re @u99of9: It would be nice to fix Z26568 but I can't spot the issue) [08:23:38] Does that mean there was a problem with the old tester? (re @Al: Changed to use Z14392 ✅) [08:26:27] Maybe another language comparison issue? (re @u99of9: Does that mean there was a problem with the old tester?) [08:33:15] The result validation fails with other equality functions (in code), and that’s because the compared objects have different representations. (re @u99of9: Does that mean there was a problem with the old tester?) [08:34:54] So does this mean we would fix it if we had type converters to code that resolved references? (re @Al: The result validation fails with other equality functions (in code), and that’s because the compared objects have different repr...) [08:45:37] Probably. But I’d be wary of converting Z11s and I think they can have different representations depending on where they come from. (re @u99of9: So does this mean we would fix it if we had type converters to code that resolved references?) [09:03:46] What are you wary about? Do we already have much code dealing with the raw objects? (re @Al: Probably. But I’d be wary of converting Z11s and I think they can have different representations depending on where they come fr...) [09:05:17] The conversion would be applied to every element in a list of labels. (re @u99of9: What are you wary about? Do we already have much code dealing with the raw objects?) [09:06:03] If a list went into code.... Yes... Why is that bad? [09:06:49] At the moment we can't really do anything useful with it in code? [09:10:23] Z24139 is pretty useful, I think! (re @u99of9: At the moment we can't really do anything useful with it in code?) [09:29:21] Did you get that to work even if the languages have different representations? (re @Al: Z24139 is pretty useful, I think!) [09:36:40] It has no failing tests but it wouldn’t work if languages were not de-referenced before execution. (re @u99of9: Did you get that to work even if the languages have different representations?) [09:38:25] Then you might support conversion to cod that just ensured de-referencing? (re @Al: It has no failing tests but it wouldn’t work if languages were not de-referenced before execution.) [09:42:17] I’m not opposed, just wary. Z11 and Z60 should work without custom conversion. If they don’t, the bug(s) should be fixed. (re @u99of9: Then you might support conversion to code that just ensured de-referencing?) [09:45:43] Agreed. Does "work" include de-referencing? (re @Al: I’m not opposed, just wary. Z11 and Z60 should work without custom conversion. If they don’t, the bug(s) should be fixed.) [09:50:42] Languages must always be comparable in code, whether from Z11, Z31, Z60… whether from Wikidata fetches or Wikifunctions objects… 🤷‍♂️ (re @u99of9: Agreed. Does "work" include de-referencing?) [09:55:34] Sounds like an excellent principle. Have we already got a phab request? (re @Al: Languages must always be comparable in code, whether from Z11, Z31, Z60… whether from Wikidata fetches or Wikifunctions objects…...) [10:14:19] I assumed the problems were fixed with T419789#11837447. Please also note T421468#11893598. But the strictly accurate answer appears to be: no. (re @u99of9: Sounds like an excellent principle. Have we already got a phab request?) [10:51:32] I think we’d currently need a new function that restricted itself to fetched Z11s (or Z6001). Should it return the empty string by default? (re @u99of9: Anyone want to try to crack a code implementation of Z35809? It would have a shot at making Z35808 the most efficient implementa...) [11:03:14] I initially had it defaulting to [mul]"" if it missed all the languages. But since the multilingual version Z34947 already had Z34949, I switched to return first on misses. I can see arguments for both. (re @Al: I think we’d currently need a new function that restricted itself to fetched Z11s (or Z6001). Should it return the empty string ...) [11:06:25] Why do we need a new function? If the current tech limits this one to fetched Z11s, that's fine. I only made it today, so it's not in use. (re @Al: I think we’d currently need a new function that restricted itself to fetched Z11s (or Z6001). Should it return the empty string ...) [11:10:29] Oh… I’ll try that, then. (re @u99of9: Why do we need a new function? If the current tech limits this one to fetched Z11s, that's fine. I only made it today, so it's n...) [13:48:00] Z35835, with an “adjustment” to Z35820. (re @u99of9: Anyone want to try to crack a code implementation of Z35809? It would have a shot at making Z35808 the most efficient implementa...) [13:52:05] Winner! Yes, it was promoted to the top https://www.wikifunctions.org/w/index.php?title=Z24766&diff=279114&oldid=279113 (re @Al: Z35835, with an “adjustment” to Z35820.) [13:53:58] for the Phab report, shall I make a disconnected version of the unadjusted test? That seems like a fairly clear case of the remaining language comparison issue. (re @Al: Z35835, with an “adjustment” to Z35820.) [13:55:41] eek, now it's out of favour, some kind of weird thrashing: https://www.wikifunctions.org/w/index.php?title=Z24766&diff=279117&oldid=279116 (re @u99of9: Winner! Yes, it was promoted to the top https://www.wikifunctions.org/w/index.php?title=Z24766&diff=279114&oldid=279113) [13:56:30] It certainly highlights that it’s a Z11 issue there 🤔 (re @u99of9: for the Phab report, shall I make a disconnected version of the unadjusted test? That seems like a fairly clear case of the rem...) [14:11:31] T427644 feel free to edit (re @Al: It certainly highlights that it’s a Z11 issue there 🤔)