I'd provide all five ci procedures; people will expect them. Case-sensitive procedures would not provide any special benefits except on a system where string comparison order is not Unicode code point order. (UTF-16 systems have that issue, but I know Kawa works around it to provide Unicode order anyway.)

On Tue, Sep 24, 2019 at 3:57 PM Lassi Kortela <xxxxxx@lassi.io> wrote:

>> Does anybody need to sort ASCII-case-folded strings?
>>
>> * ascii-ci<?
>> * ascii-string-ci<?

> Sure. The advantage of them is that they can be done by arithmetic without
> a lookup table, and thus presumably faster.

Can you list which case-insensitive procedures would be useful? Is <
enough or are <= and > and >= and = useful for something as well?

What about case-sensitive comparison procedures - is it enough to have
the Unicode ones? ASCII ones would normally give the same results as the
Unicode-aware ones, but are there some Unicode gotchas that would be
avoided by the simple ASCII versions?