L2/08-278 Date/Time: Mon Aug 4 17:31:28 CDT 2008 Contact: karl-pentzlin@acssoft.de Name: Karl Pentzlin Report Type: Public Review Issue Opt Subject: Comments on Public Review Issue #122 Comments on Public Review Issue #122 Proposal for Additional Deprecated Characters - Karl Pentzlin ( karl-pentzlin@acssoft.de ) 2008-08-04 1. Some of the "discouraged" and "additional proposed deprecations" are contained in the character collections 281 and 282 (titled MES-1 and MES-2, respectively) as specified in amendment 1 of ISO/IEC 10646:1-2000, namely: in MES-1: 2126 0149 in MES-2: 037E 0387 20A4 2126 2329 232A 0149 These collections may be referenced to in other standards, which is the case for MES-1 in the current editions of ISO/IEC 9995-2 and 9995-3. If these characters become deprecated, it is to consider what shall be happen with these character collections and other standards referring to them (nevertheless, the mentioned 9995 standards are currently in revision anyway). This opportunity can be taken to redesign these collections, especially MES-1 is defective anyway for other reasons (see e.g. http://www.csc.fi/english/pages/meek/An-alternative-to-the-current-ISO-IEC-9995-3 Appendix clause A6, p.11). 2. Some of the "discouraged" and "additional proposed deprecations" are contained in the not standardized but widely used WGL4 (Windows Glyph list 4, maintained my Microsoft, see: http://www.microsoft.com/typography/otspec/WGL4.htm namely: 0387 20A4 2126 0149 While caring about such a collection is not a duty of the Unicode consortion, Microsoft as a consortion member should be aware of it. 3. Regarding 2329 and 232A, deprecation is strongly supported due to the defective NFC mapping to characters with CJK-like behaviour. 4. Regarding 212A and 212B: As there is already a canonical decomposition to a single character which is expected to look the same anyway, deprecation is supported. 5. Regarding 037E, 0387, 2126: Due to the existing canonical decomposition to a single character, the user cannot expect that a different look intended by the font designer due to the script change is preserved as the user cannot prevent the application of the NFC in any cases. Thus, deprecation is questionable to the same extent as the canonical decomposition here, but as the latter is there, deprecation seems systematic and is therefore supported. 6. Regarding 0344, neither discourage nor deprecation seems necessary like for all other precombined characters. 7. Regarding 0149, deprecation is appropriate if the compatibility decomposition is really wrong (i.e. if 02BC is really unacceptable to Afrikaans writers). Otherwise, I see no reason to handle it in another way than the other "questionable" precomposed characters 0132,0133, 013F,0140. 8. Regarding 20A4, a discouragement or deprecation is definitively not supported. The reference glyphs of 00A3 and 20A4 have a different appearance which is clearly outlined. While the original intention for 20A4 (for lira, whichever currency [Italian, Turkish, Egypt, ...] is meant by that statement) may be wrong, it cannot be excluded that the difference is meaningful in other context (e.g. historic currencies). Without excluding this by extensive research, a deprecation (which is in fact a unification of 00A3 and 20A4) is a random action. By the way, it is proposed for 20A4 to drop the first informative note and to make the second informative note more precise by stating for which lira (Italian, etc.) 00A3 is in fact preferred. 9. Regarding 0F07...17D8, the author has no expertise. -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- (End of Report) -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- Date/Time: Mon Aug 4 19:33:53 CDT 2008 Contact: karl-pentzlin@acssoft.de Name: Karl Pentzlin Report Type: Public Review Issue Opt Subject: More Comments on Public Review Issue #122 More Comments on Public Review Issue #122 Proposal for Additional Deprecated Characters - Karl Pentzlin ( karl-pentzlin@acssoft.de ) 2008-08-04 10. If any action is done on 037E and 0387, the same should be done on (read "c.d." as "has a single character canonical decomposition to"): 0374 GREEK NUMERAL SIGN (c.d. to 03B9 MODIFIER LETTER PRIME), 1FEF GREEK VARIA (c.d. to 0060 GRAVE ACCENT) 1FFD GREEK OXIA (c.d. to 00B4 ACUTE ACCENT) as the situation is exact the same as with 037E and 0387. 11. I suggest to mark deprecated characters in code tables by "crossing out" by thin diagonal lines from the upper left to the lower right corner and from the lower left corner to the upper right corner of the whole cell, to show the casual user the deprecation by first glance. -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- (End of Report)