Ralph Johnson - Blog

If Smalltalk is the future now, why will it be the present in the future? If not now, why then? Is there something we are missing now that we will have in the future? It used to be that there was a story: when computers become fast enough, then Smalltalk, Lisp, etc. will rule the world. Well, we're there, computers are fast enough, that's not a detraction for Smalltalk anymore. And I think that's helped Smalltalk, but it certainly doesn't rule the world.

Comment on Ralph Johnson - Blog Re: What's Wrong With Smalltalk? Well, I've been spen... by Travis Griggs

Ian, the problem is that while we were waiting for computers to get fast enough, the number of programmers grew greatly and they all cut their teeth on metaphors that were grounded from the days when computers were not as fast. I do believe that computing environments are very slowly becoming more and more like Smalltalk/Lisp/etc. We've got the compute power now. It's upgrading the prevailing opinion to be comfortable with what we can now do that is the challenge.

... but it won't *be* Lisp or Smalltalk. New languages and class libraries keep borrowing ideas from Smalltalk and Lisp, and I think that's what people mean when they say that the future will look like one of our powerful but relatively unused languages today. C## looks more like Lisp or Smalltalk than does Java, for example. Maybe, far enough in the future, we will arrive at a language that is McCarthy's original Lisp, but I'd be happy just to have common languages give me a programming experience that feels enough like Lisp and Smalltalk.

Comment on Re: What's Wrong With Smalltalk? Well, I've been spen... by Rich Demers

OK, so maybe the rest of the programming world is finally, albiet slowly, catching up to Smalltalk, though they still have a lot of obsolete ideas to slough off: static declarations, file orientation, make-style recompilations, etc. But let's assume they finally get there; that they do catch up -- say in another twenty years. So where's Smalltalk at that time? Still where it is today, pushing the same set of (already) ancient ideas? Have we experimented with new ideas for the language, the libraries, and the IDE. Have we pushed into new computing territories? Or are we just content to settle the territories pioneered at Xerox Parc 25 years ago?

That's the trouble with what's going on in the curly bracket world of programming. People want to play with the language, where as Smalltalk and LISP are so simple as to allow you to do almost anything you can imagine, use any paradigm you can dream of, make any bizarre control structure you could ever want, without changing the language. Whether it be the pure messaging syntax of Smalltalk, or LISP Macros (which are even more powerful than people can imagine), you can do just about anything.

Why are people so obsessed with a new language? New does not, by default, equal better. It simply is new. Different. People built cars with all kinds of weird transmission designs, push buttons, etc., but eventually everyone figured out that one design worked better. The problem is that all the other languages are trying to to catch up with Smalltalk without realizing that they don't need to ADD features, they need to take stuff out.

...what software will be like in ten years? ... Learn Smalltalk.
Well that's vague enough to cover all manner of possibilities :-)