RE: Surrogate points

From: Jon Hanna (jon@hackcraft.net)
Date: Sat Jan 29 2005 - 12:24:30 CST

  • Next message: Hans Aberg: "Re: Surrogate points"

    > The idea was merely to straighten out the logical model, so
    > as to not be
    > confusing. That would help software development.

    No, you're suggesting we artificially create a new Y2K problem.

    There's an important design constraint faced by those developing Unicode,
    "Life sucks, get a helmet".

    It sucks that languages aren't always consistent in their use of alphabets.
    It sucks that alphabets overlap, or that languages that use a character with
    the same semantics may differ in the preferred glyph. It sucks that it's not
    always easy to determine what symbols are used inline with text or purely in
    non-textual contexts (and hence it's not easy to determine what symbols
    should be encoded). It sucks that legacy encodings sometimes made bad
    decisions and frequently made decisions that conflict with those made
    elsewhere. It really sucks that we can't just throw out all the old systems
    and start from scratch in a worldwide Year Zero and it sucks that when
    Unicode changed scope from "commercially significant" character sets to all
    characters the 65536 characters available in UCS-2 was no longer enough.

    Hey, the surrogates aren't even the most illogical part of the model by a
    long shot, the characters with single character canonical decompositions are
    much worse. They're in there because life sucks, and we the design has to
    deal with it, rather than whining about a few trivial programming tasks
    becoming necessary.

    Regards,
    Jon Hanna
    Work: <http://www.selkieweb.com/>
    Play: <http://www.hackcraft.net/>
    Chat: <irc://irc.freenode.net/selkie>



    This archive was generated by hypermail 2.1.5 : Sat Jan 29 2005 - 12:26:28 CST