[01:09:14] @jdforrester Do you have some expectations about labelling in en-generic that I don't know about? https://www.wikifunctions.org/w/index.php?title=Z21554&diff=178382&oldid=156711 (and other similar recent edits). The generic label is now inconsistent with all of the implementations and tests, and the edit overrides the original author's choice of language variant. IMO [01:09:14] we should fo [01:09:15] llow similar style guidelines to en-wiki for the generic, with alternate spellings as aliases. Then we can also freely add en-variants with appropriate spelling. Basically I have no problem with adding en-us, but don't think the changes to en were consistent or harmonious. [10:55:53] I’ll just say that it’s tricky… On balance, I’d favour consistency across the whole of Wikifunctions, for any given word or group of words. But as long as Search stems the variants, it doesn’t really matter. I did worry about “color” when it first appeared but I persuaded myself that its dominant use as a keyword should probably override any linguistic preference. F [10:55:53] [10:55:54] WIW, Google Books Ngram Viewer shows “color” as more than twice as common as “colour”, although the gap is closing. In contrast, “catalogue” is now more common than “catalog” (since 2006). The validity of any inference from such data is doubtful, of course. [10:55:55] https://books.google.com/ngrams/graph?content=color%2C+colour&year_start=1982&year_end=2022&corpus=en&smoothing=3 [10:55:57] https://books.google.com/ngrams/graph?content=catalog%2Ccatalogue&year_start=1982&year_end=2022&corpus=en&smoothing=3 (re @u99of9: @jdforrester Do you have some expectations about labelling in en-generic that I don't know about? https://www.wikifunctions.org/...) [11:21:21] [[Wikifunctions:Type proposals/Wikifunctions object reference]] is a good proposal but the name is a deal-breaker. [11:24:42] Ultimately you cannot reference every Wikifunctions object (Z1), only persistent objects (Z2), so the name is ambiguous at best and false at worst [11:30:05] I don’t think it’s ambiguous. The value is always a reference to a Wikifunctions object. It’s true that not every Object has such a reference, but I think that’s irrelevant. (re @Feeglgeef: Ultimately you cannot reference every Wikifunctions object (Z1), only persistent objects (Z2), so the name is ambiguous at best ...) [11:30:42] It is pretty relevant (re @Al: I don’t think it’s ambiguous. The value is always a reference to a Wikifunctions object. It’s true that not every Object has suc...) [11:32:49] Especially because since there are a countably infinite number of Wikifunctions objects, there *is* a way to create a reference system [11:44:03] I’m not attached to the name, so long as it’s not confused with Reference. (re @Feeglgeef: Especially because since there are a countably infinite number of Wikifunctions objects, there *is* a way to create a reference ...) [18:28:25] In https://www.wikifunctions.org/view/en/Z21121 , why does the the placeholder in the "Rounding" field say "E.g. -178"? [18:28:44] "-178" is unusually specific. Where does it come from? [18:48:27] More importantly, how do I actually run it? [18:49:11] If I put `0` in Number, `C` in From, and `F` in To, I'd expect 30, but I get "void". [18:55:46] Oh, I guess I have to fill something in Abbreviate?, Adjective?, and Language, even though they are marked as "not supported yet". [19:01:38] Does it mean that there's no way to have optional parameters? [19:25:04] It comes from Z16701, the first listed test case of Z16700, the display function connected to Z16683. (re @amire80: "-178" is unusually specific. Where does it come from?) [19:27:41] Wow. [19:28:01] From which part exactly? From the second string of "result validation"? [19:31:52] I assume it’s the actual result, but I haven’t checked that. (re @amire80: From which part exactly? From the second string of "result validation"?) [19:32:00] For the first time I pressed the Details of the result and was astonished that the Memory usage of converting 1 kg to g took 207 MiB! [19:34:16] Very roughly, 38 times that's the complete works of Shakespeare in plain text. [19:35:19] Perhaps I could write a conversion function from mb to cwoSipt (complete works of Shakespeare in plain text). [19:36:28] And... if there was no test, there would be not "e.g." at all? (re @Al: I assume it’s the actual result, but I haven’t checked that.) [19:38:44] I rather think that if there were no test, the function would not be connected as the display function (but only WMF staff have that right). (re @amire80: And... if there was no test, there would be not "e.g." at all?) [19:45:55] There is no formal optionality or default. I think only strings and lists are valid without a value (because they can be empty). A function can be created with some other type of object specified as its type, as in Z23716. (re @amire80: Does it mean that there's no way to have optional parameters?) [20:19:35] At which point does it fail? In the Python code or before that? Seeing it failing mysteriously with "void" is not very useful. If there are no optional parameters, wouldn't it make more sense, for example, not to allow pressing the "Run function" button? (re @Al: There is no formal optionality or default. I think only strings and lists are valid without a value [20:19:35] (because they can be empty)....) [20:20:19] It would make sense to me to allow this and to just set the values of those parameters to `None` in the code, or something along those lines. [20:22:10] It fails in orchestration, because the function call is not a valid object. (re @amire80: At which point does it fail? In the Python code or before that? Seeing it failing mysteriously with "void" is not very useful. I...) [20:34:10] Convert is really messy (re @Jan_ainali: For the first time I pressed the Details of the result and was astonished that the Memory usage of converting 1 kg to g took 207...) [20:35:11] Optional args are in "no current plans" (re @amire80: At which point does it fail? In the Python code or before that? Seeing it failing mysteriously with "void" is not very useful. I...) [20:35:53] With all the other epics important to the community [20:37:04] I need Wikidata property lookup before that gets done (re @amire80: Oh, I guess I have to fill something in Abbreviate?, Adjective?, and Language, even though they are marked as "not supported yet...) [20:38:56] Ah, yes… I couldn’t find a ticket for that 🤷‍♂️ (re @Feeglgeef: I need Wikidata property lookup before that gets done) [20:41:30] This is another task in "no current plans" (re @amire80: "-178" is unusually specific. Where does it come from?) [20:42:45] Is it *T382921*? (re @Feeglgeef: I need Wikidata property lookup before that gets done) [20:43:25] Not really (re @wikilinksbot: T382921 – Consider adding optimization parameters to Z6821 and other Wikidata fetch functions [open]) [20:43:36] Actually not at all [20:44:44] Which one, then? (re @Feeglgeef: Actually not at all) [20:44:59] Not sure there is one (re @Al: Which one, then?) [20:45:50] That was my original conclusion. (re @Feeglgeef: Not sure there is one)