Bill Kohl <billk at hypercon.com> wrote:
>If Smalltalk were case insensitive, then I assume the method
>'howIsThisForReadability', could be sent as HOWISTHISFORREADIBILITY or
>howisthisforreadability (plus all other combinations). For me, this would
make
>readability an issue.

Well, this does look like a really sound argument for case sensitivity.
Symbols representing selectors do come back to us as items in browser
method lists, or in "messages" selection dialogs.  All technicalities and
personal preferences aside, this is an actually important reason to
preserve their original shape.

<jarvisb at timken.com> wrote:
>I personally prefer case-sensitive languages to case-insensitive
>languages.  If I want to have selectors or procedures named abc, Abc,
>ABC, and aBc in a program I darn well expect the language to allow it
>and not try to limit what can be done.  I also *HATE* underscores and
>dashes in identifier, selector, or procedure names.  If I wanted to
>PROGRAM-IN-COBOL or PROGRAM_IN_PL_1 <gag!> I'd do so.  I don't.  Keep
>Smalltalk case-sensitive!!!!  :-)

Hey, dashes are just fine to separate words!  But underscores -- *those*
are really ugly, I'm with you on that.  Seriously, as long as there are
several approaches to achieving the same result, there are that many camps
religiously defending the ergonomical, historical, or moral merits of their
favourite one.  In a well-known primitive case, the only problem was
choosing the proper end of an egg, while in our advanced world
opportunities are definitely much richer.  I believe, Minsky described
something similar in his _Society of Mind_: the closer two alternatives are
to each other, the harder it is to choose -- the paradox being that the
closer they are, the less significant is the choice.

-- 
Vassili Bykov         vassili at objectpeople.com
The Object People     http://www.objectpeople.com
+1(613)225-8812