On 4/20/07, Glenn Linderman <perl@nevcal.com> wrote: > "codepoint string" could be a reasonable name. and with a new name comes the opportunity for a new implementation. codepoint string scalars might not be "just" utf8, but might be utf16, or some monster made with linked lists of COW elements, facilitating s/// operations on large codepoint strings with less memory used. Carry on. :) -- Is the opposite of "tax and spend" "borrow and burrow?"