Remove the need for #asByteString

[ghost]

I'm just starting to learn Smalltalk, Cuis, and reading this https://cuis-smalltalk.github.io/TheCuisBook/Message-to-string-entities.html . I find having to use #asBytestring so often is a nuisance. Whether the (new) strings are implemented with 1 byte, or 1..4 bytes, per character is internal implementation detail that should not concern the end users of Cuis. It's like exposing users how intergers or others are implemented in Cuis.

IMO, the new strings should be able to correctly receive #at and #indexOf , as well as be changed using #at: #put: etc, without having to be converted to byte strings first (or at least not by the user, explicitely).

Also: I enter the array literal:

($h $i) .

and Cuis responds with: #($h(000068) $i(000069))

Or else, it goes against the simplicity and cleanliness aims of Cuis/Smalltalk, no?..

[jvuletich]

I agree that all Strings should be fully polymorphic, and users, especially people learning about Cuis, should not bother with their differences, or need to know about #asByteString. Still, UnicodeString doesn't support #at:put:. I think that in general modification of String should be discouraged, and not taught in an introductory text. I also suggest discussing this at our mail list https://lists.cuis.st/mailman/listinfo/cuis-dev . That's our forum for public discussion.

[ghost]

Alright, thanks for replying; I will write on that mailing list such suggestions in future.

[jvuletich]

This has already been addressed in The Cuis book.