On Sat, Apr 2, 2016 at 5:39 AM, John Cowan <xxxxxx@mercury.ccil.org> wrote:
> Arthur A. Gleckler scripsit:
>
>> John, how's the implementation coming?  Do you think we'll
>> be ready for last call soon?
>
> I haven't started, but there is little to do.  Basically I need to
> take Olin's implementation, remove any unneeded procedures, add the
> trivial string-cursor-* and string->vector/cursors procedures, and write
> string-split.  That will produce an implementation in which cursors and
> indexes are the same, which satisfies the requirements of the SRFI.

Given the nature of the SRFI I don't think an implementation
without disjoint cursors is sufficient.  I'll write one for Chibi.

I'm slightly concerned about the feasibility of implementing this
efficiently in other implementations.  In Chicken with the utf8 egg
at least it would require heap allocation which could make some
string algorithms measurably slower.  I'd also like to hear from
implementors of other utf8/utf16 Schemes (gauche, kawa, etc.).

[I like the new draft very much, btw, I just haven't had time to
review in detail yet].

--
Alex