> -----Original Message-----
> From: Doug Way [mailto:dway at mat.net]
> Sent: Wednesday, November 17, 1999 12:54 PM
> To: 'squeak at cs.uiuc.edu'
> Subject: Command Line Workspace (was RE: Language implementation in
> Squeak)
> 
> 
> 
>> > 
> > Well, I guess I knew there were a lot of people who prefer 
> workspaces.
> > I just haven't made the leap yet... alas. 
> 
> I assume by "listener" you mean a command line interface
> (interpreter/evaluator)?  (Sorry, I've only done a little 
> Lisping, and may
> not be up on the terminology.)
> 
> Given a choice between using a straight command line and a 
> workspace, I'd
> take the workspace.  But there have been times when a command 
> line would
> be quicker to use, such as evaluating a bunch of different short
> expressions one after another.  (With a history mechanism of 
> course... a
> command line without a history is worthless.)
> 

Absolutely.

> It seems like it wouldn't be too hard to create a hybrid command line
> interface/workspace in Squeak.  Call it a "command line workspace".
> Hitting <return> after typing an expression would simply do a 
> "print it"
> on the expression, output on the next line.  And you could 
> still have the
> full text editor menu with "do it", "inspect it", etc., if 
> you wanted to
> highlight text and do workspace-like stuff.
> 

That would be great... It would be very similar to using a lisp 
interpreter within emacs (or an IDE like LispWorks) which 
works very well in my experience.

> Although if you wanted to enter multiple lines of code (and 
> be allowed to
> include returns in it) for a single evaluation, you would 
> need a mechanism
> to do that.
> 

Sure.  Say, maybe you could start the expression with an open
parenthesis and evaluate it when you see a return
following the matching right parenthesis :-)