At the risk of sounding like an enemy of the proletariat, an unreconstructed
reactionary, *and* an AOL-bot, I agree (yet again)!  Having had to deal with
APL on and off I have to say that the prospect of unlimited use of
unstandardized symbolic binary messages fills me with...well, heartburn is
probably the nicest way to say it.  I think that a balance should be drawn
between the conciseness of symbolic binary messages, and the readability of
keyword messages.  My choice would be, as is the case now, to set things up
in such a way that keyword messages are favored.  In my opinion the use of
keyword messages forces us to think about what we're trying to communicate
more than the use of symbolic messages does/would, except in those limited
cases where the symbolic message (e.g. #+, #-, etc) are deeply embedded in
our cultural biases.  (OK, this begs the question of such things as #-> for
creating an Association, #@ for creating Points, etc).  Today Squeak code is
pretty easy to read once you get past the initial "it ain't C" reaction.  I
would find it distressing to have this "readability factor" compromised.

Bob Jarvis
The Timken Company

> -----Original Message-----
> From:	Peter William Lount [SMTP:peter at smalltalk.org]
> Sent:	Tuesday, March 16, 1999 6:21 PM
> To:	squeak at cs.uiuc.edu
> Subject:	Re: Why so few binary method selectors? Because they're bad!
> 
> Hi,
> 
> Readability is very important for Smalltalk code. If we're not carefull it
> will end up looking like apl code! Let's not get to crazy with those
> #&*%#$*#&%@*!$& binary selectors!
> 
> sameObjectAs: would be better than sameAs: or ==.
> 
> Peter William Lount
> 
> ----------
> From: Reinier van Loon <R.L.J.M.W.van.Loon at inter.nl.net>
> To: squeak at cs.uiuc.edu
> Subject: Re: Why so few binary method selectors? Because they're bad!
> Date: March 16, 1999 12:40 PM
> 
> >down, let me ask the group a question. If we could have long binary
> >method selectors, what would be the first one you would define, and what
> >would it do? Best regards. -- Ward
> None.
> 
> I think binary selectors make code harder to read.
> I'm still having troubles explaining == and = to people.
> sameAs: and equals: would have been better IMHO.
> 
> The other problem with binary selectors is that they might get multiple
> meanings.
> 
> Reinier.