2025-04-07 [06:16:53.0316] I looked for the `islamic` and `islamic-rgsa` calendars in HTTP Archive. Most matches were bundlers including a chunk of CLDR data. Then false positives (not calendar) for "islamic" in quotes. Finally, I narrowed the list to 6 URLs that might actually be about relevant usage. None show `islamic-rgsa`. All 6 are either under .sa or .ae. Of these, 5 seem to format a date using `islamic`. I can't tell if these pages get the results they want, because today `islamic` agrees with `islamic-umalqura`. Need to re-test on a day when they don't. Anyway, I'm pretty sure at least the .sa pages would be better off using `islamic-umalqura`. [06:17:15.0900] CC eemeli dminor sffc [06:19:55.0959] CLDR says .ae prefers Umm al-Qura, too. [06:21:50.0277] Interestingly, mapping the calender by region might make more of a mess, since I see pages under .sa use ar-FR-u-ca-islamic and ar-TN-u-ca-islamic. No idea why. [06:29:10.0359] Also, earlier today, I opened https://github.com/mdn/content/pull/39012 2025-04-08 [01:14:41.0887] FWIW, ar-FR is not a thing in CLDR, but ar-TN is. Furthermore, ar-TN affects how month names are formatted in some cases. 2025-04-09 [03:42:17.0962] I filed https://github.com/mdn/content/issues/39050 . 2025-04-13 [16:14:29.0065] USA is down for 5m 402 report tomorrow. lmk if delta 2025-04-14 [18:04:15.0467] sir, you can't just say things like "USA is down" in these times [06:54:08.0325] I think he meant the shitcoin, which is indeed down: https://www.coingecko.com/en/coins/american-coin 2025-04-16 [00:02:41.0164] Here are my slides on the non-binding Intl Era Month Code Stage 2 Update: https://docs.google.com/presentation/d/1wvJoRFa8nRjlYSHuVLpxx-wCfwt4H9NIw2fsGJ72gxs/edit#slide=id.p (CC hsivonen) [00:59:52.0779] Thanks. I've noted slides 8 and 9 for Gecko patch review purposes. [03:30:51.0257] sffc: On slide 9 you have `new Intl.DateTimeFormat('und-SA', ...)`, but AFAIK no runtime does anything with the region there. Is that intentional? [03:33:50.0722] Though `Intl.Locale` does maximise that to `ar-Arab-SA`. For a locale tag like that, that does feel like it'd be closer to what the user wanted in this case. [03:40:12.0527] Also, "Anno Minguo" is pretty apt. [04:59:54.0026] Actually, should we consider a new `localeMatcher` option value that did apply maximization to the selected locale? Or is that something that's technically already allowed with `"best fit"` locale matching? [06:00:29.0398] It is intended for locale extension resolution, i.e. mapping `u-ca-islamic` to `u-ca-islamic-umalqura`. (We should make sure the spec allows this) [06:00:56.0822] * It is intended for locale extension resolution, i.e. mapping `u-ca-islamic` to `u-ca-islamic-umalqura`. Some other region might map to `islamic-tbla` or `islamic-civil`. (We should make sure the spec allows this) [06:01:15.0638] * It is intended for locale extension resolution, i.e. mapping `u-ca-islamic` to `u-ca-islamic-umalqura`. Some other region might map to `islamic-tbla` or `islamic-civil`, but this is an open question. (We should make sure the spec allows this) [06:03:45.0221] I read that some countries are working on launching a satellite to observe the crescent, and presumably after that happens, there might be another reliably predictable islamic calendar, say `islamic-satellit`, that some regions might start using. [06:20:43.0015] I'm pretty sure that the spec does not currently support that, unless an implementation specifically includes `und-SA` in _availableLocales_, or with "best fit" matching matches that to something like `ar-SA`. [06:21:30.0756] Acknowledged [06:23:22.0683] I'll change the slide to ar-SA to avoid this question during the presentation 2025-04-18 [10:30:23.0710] terminology question, is "language tag" or "locale id" more common? i see both IsStructurallyValidLanguageTag and CanonicalizeUnicodeLocaleId AOs [10:30:31.0119] that is, assuming they are the same thing. perhaps they're not? [10:56:07.0203] https://tc39.es/ecma402/#sec-language-tags > This specification identifies locales using *Unicode BCP 47 locale identifiers* as defined by [Unicode Technical Standard #35 Part 1 Core, Section 3.3 BCP 47 Conformance](https://unicode.org/reports/tr35/#BCP_47_Conformance), and its algorithms refer to Unicode locale nonterminals defined in the grammars of [Section 3 Unicode Language and Locale Identifiers](https://unicode.org/reports/tr35/#Unicode_Language_and_Locale_Identifiers). Each such identifier can also be referred to as a *language tag*, and is in fact a valid language tag as that term is used in [BCP 47](https://www.rfc-editor.org/rfc/bcp/bcp47.txt). [10:56:28.0621] "locale identifier" is more formal [10:59:39.0474] thanks