🚀 go-pugleaf

In article <19921101.003@erik.naggum.no> enag@ifi.uio.no (Erik Naggum) writes:
>F Avery Bishop <fab@fungus.zso.dec.com> writes:
>|
>|   Unicode:
>|      Unicode is an international character encoding standard that
>|      encompasses all the world's national scripts in a 16bit code
>|      space.  Unicode is a profile of the International Standard ISO
>|      10646.  Supported by most major computer and software vendors,
>|      Unicode greatly facilitates the development of internationally
>|      accepted software.  All currently used character sets can be
>|      transferred to Unicode.
>
>I'd like to see this last statement supported by some evidence.  I have
>studied IBM's Character Data Representation Architecture, including the
>Registry and the Graphic Character Global IDs, and IBM's registry of
>code pages, and I have found several characters that are not included in
>Unicode.  I have studied the ISO Registry of Coded Character Sets to be
>Used with Escape Sequences (maintained by ECMA), and many bibliographic
>character sets include characters not found in Unicode.  There is a
>character in Apple's Greek character set that is not in Unicode.
>(Thanks to Michael Everson for this last finding.)  Also, several
>_control_ character sets cannot be transferred to Unicode at all,
>because it relies on ISO 6429 for control functions.
>

  First, let me point out that I didn't write the above -- I posted an
  announcement that someone else had written.  Having said that, let
  me now defend the statement.  It certainly is the intent that all
  major national and corporate character codes be mappable to Unicode.
  IBM, Apple, Microsoft, Digital, RLG (representing bibliographers)
  and other vendors with their own character encodings where heavily
  involved in the design of Unicode to insure that 1) national
  standard character codes could be mapped to Unicode and back, and 2)
  other widely used standards could be mapped where feasible.

  A couple of notes about claims of characters in other standards
  that are not covered in Unicode:

  1) Some characters are covered in unicode as a combination of a base
     character and a non-spacing mark, (also called 'combining marks',
     formerly called 'non-spacing diacritics.')  Examples are certain
     rare Hangul blocks in Korean (although these aren't in any
     standard today anyway), some characters in scripts of North
     American natives, IPA characters, many combinations used by
     bibliographers, etc.

  2) We often hear claims that some characters aren't covered, but
     find that they are when given the specifics.  If you would
     provide the code points (or better still, FAX me a typical glyph
     rendering), I'll follow up and find out for sure whether or not
     they are covered.  My FAX number is +1-206-865-8800.

  3) Many characters are assigned code points in a Compatibility Zone
     (U+FE00 through U+FFEF) and a CJK Compatibility Zone (U+F800
     through FAFF).  These would not have been encoded, generally
     because they conflict with Unicode principals, but were included
     for compatibility with existing national and industry standards.
     Many of the IBM corporate characters are in these zones.

  4) A few scripts were intentionally delayed or withdrawn either
     because of lack of information, or to promote the 10646 merger.
     These include Tibetan, Burmese, Inuktitut/Cree, Sinhala, Khmer,
     Mongolian, Ethiopic, and dead scripts of interest to scholars
     such as Egyptian Hieroglphics.  However, none (-- In my
     understanding) of these scripts has a standard encoding yet, so
     the compatibility is not an issue.

  5) Rendering variants are often cited as characters that are not
     covered by Unicode.  As a CHARACTER code, it is not within the
     scope of Unicode to include glyphic variants, as a glyph registry
     would.  In particular, rich text often contains variants whose
     underlying character code is identical to another that appears as
     a different shape.  The fact that not all (of the continuously
     infinite number of) glyphic images are displayed for a particular
     character in Unicode has led some investigators to conclude that
     their favorite 'character' is not included.  This principle of
     parsimony in assigning code points only to characters (with
     exceptions for compatibility) is applied not only in in obvious
     cases such as italic and bold, but also in rendering variants in
     Arabic and Indic languages due to word position.

  If you really have found some characters
  that should be then we certainly want to know about that so we can
  include them in Unicode 1.1.

>I'm not qualified to state that these characters are widely "used," but
>I find it strange that others haven't arrested the over-active Unicode
>people on this point before.  I'm tempted to interpret the statement as
>saying that "used" means "used by vendors supporting Unicode."
>Information, please, and less marketing.
>

  Well, I have to agree that much of what passes for information in
  this and other fora (isn't that the plural of 'forum'?) is either
  rhetoric from militants frothing at the mouth with passion for their
  own cause and disdain for all else, or hype that is more fluff than
  substance.  However, I believe Unicode has been reasonable in its
  public release of information.

  Unicode has felt the need to do some marketing because it was the
  underdog in the world of standards until a couple of years ago.  At
  that time, Unicode was still a little known group of innovative
  linguists/hackers who were regarded by the establishment of
  standards professionals as eccentric at best, and dangerously
  incendiary at worst.  Unicode is making an effort to provide
  accurate information on its activities in a professional manner.
  Call that marketing if you like.  What's wrong with that?
  
/fab
-- 
F. Avery Bishop, (206)-865-8916 | I don't usually do this, but this one deserves
Digital Equipment Corporation   | the standard disclaimer -- These opinions are
fab@zso.dec.com	                | my own, not necessarily those of Digital 
Bellevue, WA                    | Equipment Corporation or Unicode.