Re: [idn] call for comments for REORDERING

["Soobok Lee" <lsb@postel.co.kr>]

----- Original Message ----- 
From: "James Seng/Personal" <jseng@pobox.org.sg>
To: "Soobok Lee" <lsb@postel.co.kr>; <idn@ops.ietf.org>; "Martin Duerst" <duerst@w3.org>
Sent: Friday, October 19, 2001 1:14 PM
Subject: Re: [idn] call for comments for REORDERING


> Your answer does not answer how we going to update re-ordering in future
> with new codepoints. Do you expect to continue to work with this and to
> have update RFC for it? And we going to update RFC again and again? The
> typical solution IETF use for this is IANA action and that involves a
> well-defined policy which IANA could reused for future addition. But if
> we do so, we also need to consider if we are imposing codepoints issues
> which is best dealt with other organisation.
> 
> As I repeated, I really like to point to another organisation for these
> tables if possible. If ISO 14651 is not suitable (as Ken & Mark said),
> then we look at some others. If there is none, then maybe we should
> consider NOT to do it at all.

James,
I full-heartedly welcome any change to REORDERING tables from authoritative
organizations. 

But i dont' agree on using ISO 14651  like Mark and Ken.

 REORDERING has no intersection with ISO 14651 (string comparison service), which is about collation of unicode points, how to sort unicode strings.
ISO 14651 has nothing to do with what REORDERING would provide with.

REORDERING exists for reducing successive code distances in ACE encodings.
REOREDERING is based on RENUMBERING code points values by character frequency, 
but ISO14651 takes some variants of lexicographical ordering for various needs
and exists for comparisons , not for supporing subtraction  between
any two code points values.

ISO 14651 is totally irrelvant to REORDERING.

REORDERING tables are hidden in the ACE routines, not for reference 
material for the general public. 
The reordering mappings are a kind of huge tuning paramters set geared for 
ACE.

Tuning parameters need authorities ?  NOT bad idea, but also not neccesary
because frequency based tables are always suboptimal in its nature  however
authoratitive organizations approve and issuse them.


Soobok Lee