Default locale should be en, not en-US

[getsnoopy]

It seems like WordPress currently has a unnamed locale, which is the default when no translation packages are installed, but it’s actually used as en-US. The problem with this is that almost none of the world uses any of the conventions used in it. For example, the world uses Oxford English on the international stage (the UN, ISO, the IEC, the BIPM, and many people) and British English (and its variants) in countries like the UK, Australia, New Zealand, South Africa, etc. US English is only used in the US and some historical colonies / affiliated countries such as Liberia, Philippines, and South Korea.

Furthermore, the regional settings which are used worldwide are almost invariably ISO 8601 or DMY for numeric dates (YYYY-MM-DD or DD/MM/YYYY, respectively) and RFC 2822 for full text dates (DD Month YYYY). Basically no one other than the US uses the MDY date formats (MM/DD/YYYY and Month DD, YYYY).

Given that WordPress also specifically tags pages with full locales (i.e., with regions; e.g., en-US) in the lang attribute on the root element, it’s clear that this should be changed. I’ve seen countless websites which probably were quickly set up without thinking about the locale, so they’re tagged with en-US when all of their content is in en-GB or some of the other more international variants.

My proposal is to copy all the strings from the code base (which are presumed to be in the en-US locale) to an actual en-US translation file, and change all the default text / formats to international ones such as Oxford English, ISO 8601, and RFC 2822 under the locale en, which would be labelled English (International) or simply English.

I've floated this idea in the #polyglots Slack channel, and many agree that it should be done, so I want to propose it formally.

[casiepa]

I'm all in favor of creating an en_US locale. Great start.
Modifications will be needed on Glotpress and translate.wordpress.org to accommodate the new locale and indicating a different 'source' language, which is now considered en_US

It's still pretty early here and I only had 1 coffee yet, but I'm not so sure of why 'English (int)' needs to be created. Should it not just be called 'source language'? If I create a plugin and the basic strings are in French or Italian, then that would be considered 'English (int)', which is still incorrect.

More heads will be needed here to dig into it, but thanks a lot for opening the ticket!

[ocean90]

Changing the default locale isn't quite an option as it would impact so much without a real benefit.

There are already a few English locales you can choose from, namely: en_AU for Australia, en_CA for Canada, en_NZ for New Zealand, en_ZA for South Africa or en_GB for UK. Or art_xpirate for English Pirate – once finished. There's also a ​pending request for India.
Not sure what a request for English (International) would look like. Might be worth to discuss this with the existing locale teams of the English variants on ​make/polyglots.

The value of the language attribute can already be filtered with the ​bloginfo filter for the language key or by filtering the translation for the html_lang_attribute string.
We could consider returning en by default if the locale is set to en_US but should probably do this only for new sites.

[SergeyBiryukov]

Hi there, welcome to WordPress Trac! Thanks for the ticket.

We're already tracking this feature request in #48152, let's continue the discussion there.