Characters

[David A. Moon <Moon@STONY-BROOK.SCRC.Symbolics.COM>]

    Date: Thu, 27 Jul 89 11:53:49 PDT
    From: Richard P. Gabriel <rpg@lucid.com>

    I thought the notion of ``repetoire'' was replaced by ``script.''

I have fairly good records on this character stuff.  It was "registry"
that was replaced by "script" (in March).  "Registry" was then reintroduced
with a different meaning.
					   
"Repertoire" is just a synonym for "any [mathematical] set of characters."

    I just saw this in the draft:

    ``{\datatype Standard-chars\/} are a subrepertoire of {\datatype
    base-characters}.''

    Should I replace it with

    ``{\datatype Standard-chars\/} are a {\word sub-script\/} of {\datatype
    base-characters}.''

    since ``subscript'' is wrong too?

Subtype is the right word here.  Note that in June we eliminated the
distinction between named repertoires and types, since types are also
named sets of objects.  A named character repertoire is just a subtype
of CHARACTER.

Be extremely extra sure while editing in this area not to introduce any
new "registry" versus "repertoire" typos.

    Date: Thu, 27 Jul 89 14:02:45 PDT
    From: <masinter@arisia.xerox.com>

    I'm not clear what part of the text of the character proposal was
    actually passed. 

I have fairly good records on this character stuff, but only on paper,
and my handwriting isn't good enough to optically scan them in.  Linden
promised to update the character committee document immediately after 
the meeting, but I have not seen any sign that he has done so.  Should
we harass him?  Or find somebody else to do it?  I really think the
final version of that document is valuable for archival purposes, and
should be an X3J13 document.  Even the stuff that did not get included
in the language this time around is valuable as a possible framework
for the future and should not just be lost.

    The reason why character terminology is difficult is because the words
    refer to things that people use in their every day life and outside of
    LISP. I think the words "repertoire" and "script" are used to explain
    the 'real world' foundations of actual use of characters in written
    language and actual use of characters in ISO standards for the
    interchange of information containing written language.

    Here's my opinion:

    When we want to talk about Common Lisp constructs, we can use Common
    Lisp terminology, and leave the "repertoire" and "script" to our
    explaination of how the Common Lisp terminology maps to the real world.

"Repertoire", "coded character set", and "character script" -are- Common
Lisp terminology, now that character issue 2.0.1 has passed.  Of course
this doesn't mean we have to use those terms to the exclusion of existing
Common Lisp terms.  In particular I can think of no case where the
word "repertoire" needs to be used in the Common Lisp specification.

    Thus, you should replace it with "STANDARD-CHAR is a subtype of
    BASE-CHARACTER." As in,

Agreed.

    "STANDARD-CHAR" is a type, that corresponds to a script.

STANDARD-CHAR is -not- a script.  See the definition of script in
the character committee report.

    STANDARD-CHAR is a subtype of BASE-CHARACTER. BASE-CHARACTER also
    corresponds to a script, 

BASE-CHARACTER is certainly not a script, since its definition is
completely implementation-dependent except that it is a supertype of
STANDARD-CHAR.

			     and the characters that correspond to
    STANDARD-CHAR are a subset of those that correspond to BASE-CHARACTER.

True.