Dan Oscarsson wrote:
> Applications only having access to the ACE form of names will only be > able to match names using the ACE form. They may be decoding the ACE sequence and using the canonical UCS characters, and only use ACE for transfer purposes, possibly because the protocols have not been upgraded, or because negotiation has failed. In those situations, the accessible range of the unencoded UCS characters will be restricted to lower case. > When I start using non-ASCII in my host names they resolver routines > will return native names, not ACE, to the applications. Correct. However, if the names were encoded in ACE for transfer purposes, then only the lowercase characters will be available. -- Eric A. Hall http://www.ehsco.com/ Internet Core Protocols http://www.oreilly.com/catalog/coreprot/
