Strings

[Simn]

No description provided.

[Simn]

Could you tell us how far you got with this?

[ousado]

As far as Unicode is concerned, it should be a straight mapping to the ICU API for the most part. Haven't done it yet.

We need to decide on how we go about strings in general. From what I recall from our conversations on IRC, we postponed that decision. The options were along the lines of having 2+ genc versions, one using neko-style raw, unchecked strings and one using ICU strings and hence UTF-16 internally for the default String type. We also considered another version using UTF-8 internally, which is something ICU also provides limited (compared to UTF-16) support for.
Is that approach too complicated/confusing for the user, prone to introducing incompatibilities?

[Simn]

We could start with the ICU version and then look into optimizing it I suppose.