Make sure future localization story is sane

[brson]

Localization: it's super-important for a big chunk of the world, and most projects do it poorly. We're not going to try very hard for 1.0 but I want to at least think about it and make sure we have a plausible story for post-1.0. Nominating.

[pnkfelix]

Assigning 1.0, P-backcompat-libs.

(Its not clear how broad the requirements are, I think we agree that format! is a part of it, but not clear how far it goes beyond that.)

[zbraniecki]

Hi guys, Mozilla l10n-drivers teams is interested in helping with the localization story for Rust (and later Servo).

Our approach would be to provide full l10n API similar to what we're using in FirefoxOS or L20n projects.

[alexcrichton]

Nominating for removal from milestone. We talked with @zbraniecki and I believe we have concluded that we're in a solid place to tackle this post-1.0.

[brson]

Some detail: Moz is developing https://wiki.mozilla.org/L20n to handle Firefox's localization needs, and hoping to standardize this API in JS. Localization is hard, and serious software tends not to use the localization support from any language platform. It's best if we make no effort to solve this problem badly and leave it to a completely library-based solution that can be evolved in the community and officially endorsed once it's fully-baked.

What we do need to investigate still is internationalization and locales. Internationalization is the mechanical conversion of locale-specific types like dates according to standard rules, whereas localization is the human translation of locale-specific data such as strings.

i18n is the subject of #14494

[brson]

This is continued in #14494 and #14495. Closing.