HTML should have a normative reference mandating ECMA 402

[littledan]

We should also consider having HTML make decisions on places that ECMA 402 leaves up to implementers, such as tc39/ecma402#113 and tc39/ecma402#111 , though personally I think those would be better solved at the 402 level long-term.

[annevk]

We can just make it part of the existing JavaScript section/reference. (PR welcome!)

As for the variability, are implementations actually going to expose hooks for hosts or is this largely theoretical purity? Does 402 expose hooks for hosts today?

[littledan]

There are no hooks on the variability. You can see some arguments for motivation at tc39/ecma402#111 (comment) and tc39/ecma402#110 (comment)

[littledan]

In current web browsers, the only source of variability is different decisions by implementers, not, e.g., different data files being present.

[annevk]

Thanks, left my point of view there. It sounds like we should fix the dependency as part of this issue and leave host integration for later if actual hooks for hosts emerge.