The fact that even bad Smalltalk "reads" fairly well is probably the thing I like best about it. The project I'm on now has the normal share of bad objects (any large project will have them) - but even the bad code can still be parsed by the mark one eyeball better than just about anything else out there. As David Nolen puts it:
Smalltalk has so very few concepts - it's truly stunning. There's nothing more powerful in aiding readability than a small core set of concepts. In this sense I think Smalltalk continues to be one of the few languages to get anywhere near LISP. Most languages these days are just an abomination of features trapped inside of a compiler.
The way I like to explain that to people is this: The entire syntax of Smalltalk can fit on an index card. How many pages does it take to explain the syntax of, say, Java?
Technorati Tags: development