Stephane Bortzmeyer writes:
On Wed, Feb 27, 2008 at 10:26:44AM -0800,
Mark Crispin <mrc(_at_)Washington(_dot_)EDU> wrote a message of 162 lines
The actual correct collation, assuming(!) surname-first collation
and Latin character ordering(!!), is:
due to where the surname is located in various cultures.
Is it a good idea to sort on the ordering of the sender's culture? If
the ordering is to be useful for the human user, it should be
according the receiver's culture, no?
There's support for that. You have to read a few drafts and RFCs, though.
The draft in question defines a sort command and some sort keys, and it
permits defining more sort keys. Other documents allow defining
collations (sort orders) and using them im IMAP.
Sorting by From field the way most MUAs do requires defining a new
collation and a new sort key, since this draft's "from" key uses only
the localpart ('arnt' in the case of this message). Personally I think
that sort key isn't at all fortunate, but it's too late to change that.
There's decade-old running code already and the most important thing is
to document the deployed extension.
IETF mailing list