Dan Oscarsson writes: > I have had case insensitivity in my software from the beginning. It is > not hard work. Very easy.
We're not talking about folding A-Z into a-z. We're talking about case insensitivity for _Unicode_. Case insensitivity for _internationalized_ domain names requires quite a bit of extra code in many programs. Consider, for example, a system administrator telling his MTA ``accept mail for pi.cr.yp.to'' with a Greek pi. Case insensitivity for IDNs means that the MTA also has to accept mail for Pi.cr.yp.to. (Of course, Costello is still denying that MTAs, DNS servers, etc. are affected by IDNA at all, because Costello refuses to acknowledge that these programs have user interfaces.) Interoperability requires that we continue folding A-Z into a-z inside domain names. But it doesn't require insensitivity for non-ASCII names. > When you speak > the case of letters are not heard so the user do not know what to type. ``Capital T, lowercase e, capital X.'' Typical users avoid all these hassles by consistently using lowercase, contrary to the claims made by certain people on this mailing list. For example, someone here said that we needed uppercase because the World Health Organization was ``much easier to recognize'' as WHO.org than as who.org; but the WHO web site consistently uses lowercase who.int. Anyway, (1) your notion of speakability is contrary to the entire concept of IDNs, and (2) I don't see you telling the standardization groups for UNIX, URLs, French, etc. that you demand insensitivity. ---Dan
