At 10:29 10/02/2012, Yoshiro YONEYA wrote:
Dear all,
Nemoto-san (new co-author) and I submitted new version of precis mappings
document. Please read and give your comments/suggestions.
It seems that two types of mapping are missing depending on the way
you want to address orthotypography. As a general matter I am
embrassed by the terminology confusion between the proposed term of
(string) "classes" (a term not used in the Charter) and (DNS) "CLASSes".
FYI, ICANN started advising open-roots to use an alternative CLASS
(ICP-3, 2001). I personnaly use CLASS 65300 (UFEC) for IDNA2008 testing.
I suggest the following additions:
1. the DNS Mapping Mapping, i.e. the Mapping table to map with the
applying DNS CLASS.
IDNA2008 applies to IDNs in every DNS CLASSes. Therefore, the support
IDNA2008 in different application protocols calls for the
simultaenous cross support of the DNS network application protocol.
It seems that such a support might be implemented through a mapping
table set per CLASS and in defining an appropriate DNS resolver
information tool.
2. the language Mapping.
IDNA2008 addresses the end to end support of IDNs and is therefore
only dependent on the lowercased script oriented result of the
punycode algorithm. Mapping is necessary because the information
entered by the user and possibly necessary to the protocol is fringe
to fringe. There are several ways to support the IDNA2008 unused
information: through the environment, metadata, an "punyplus" fringe
to fringe end to end transparent extension of punycode.
jfc
_______________________________________________
precis mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/precis
