Universal Multiple-Octet Coded Character Set  (UCS)


                                              ISO/IEC JTC1/SC2/WG2 N 2294

                                                                                    Date:  2000-10-27


TITLE:                        Feedback from Japan on N2271 – Proposal by TCA to amend
two source code changes in BMP CJK Unified Ideographs block

SOURCE:                   Japan, T. K. Sato

STATUS:                    Input to proposed AMD1 – 10646-1:2000

ACTION:                     For review by WG2

DISTRIBUTION:    Members of JTC1/SC2/WG2



This contribution paper proposes the correction procedure of the error described in WG2 N2271

“Proposed to amend two source code change in BMP CJK Unified Ideograph block (TCA)”.


The requirement from TCA is reasonable and it is necessary to correct the error described in the N2271.

However, to avoid the confusion of naïve users, simple correction approach should not be taken.



Because the characters in error is so similar for naïve user, the pulished ISO/IEC 10646-1 is delivering a message of  “those two are to be UNIFIED”.  Thus, it is necessary to state “those two are not unified any more”.


Proposed solution.

1.      Publish the corrigendum for ISO/IEC 10646-1:2000.

2.      The reason for the corrigendom should NOT be simple relocation of the sources.

Such as simple change of:

5B90 T source to T4-225B, and make T source of  2-19F7 to T4-2625 should not be done

And also move T2-662B from 7921 to 7934 only should not be done.


(but, actual result of the corrigendum should be so)

3.      State clearly on the corrigendum as follow:

3-1.      Disunify T source form 5B90 and 7921.   The reason of disunification should be “mistake”.

3-2.      Add source T4-225B for U-5B90,

Add source T2-662B for U-7934    then

3-4         Merging source T4-2625 to Part-2-19F7 does not requie formal process.

Just sneak it in to ext-B (and superCJK) while its development process.


I hope if corrigendum describes above, it can provide a message that those are not to be unified clearly to the user and may prevent the misuse of the UCS.