Learning new tricks

[joflasher]

One of my managers, Ric, is a SmallTalk evangalist. He can't really help it. He's an old school algorithms developer who never really got C, Prolog or Lisp. He doesn't like them and can't really use them effectively. Too much mucking around in the weeds, according to him. An ideas man with no real testbed. SmallTalk is the perfect language for him. He's very enthusiastic about introducing it to anyone and everyone who'll listen. He's been trying to convert me for years.

Now, he has a charge number and a project to enforce the indoctrination. See, SmallTalk is not production acceptable. Since Ric is doing all of his prototyping in it, we need translators to get his work into C or Java for everyone else to use. That's where I come in. Since UAV BattleLab has been gut shot, I'm short a project, I work well with Ric and I have a reputation for picking up new languages quickly. So off to the SmallTalk tutorials I go.

I feel like a biologist upon discovering a new slug. I can admire it for its form and function. I can admit that some parts of it are even strangly elegant. Still, my overall reaction is a slightly squicked "eh!".

Comments

[tg2k.livejournal.com]

How would you compare SmallTalk to Java or C#? (To the extent that you know these languages)

[msde.livejournal.com]

For some reason, I always had SmallTalk as being categorized as similar to Lisp. Well, apart from the syntax.

[prusik.livejournal.com]

I've never done anything non-trivial in SmallTalk but I've always thought it was a pretty neat language (bound up in a cool but production-wise impractical development environment).

I'm also a fan of its C-based cousin: Objective-C.

The odd thing is that this is despite the fact that I'm not a particularly big fan of object oriented programming (at least as practiced by C++ or Java). I think all languages aspire to become usable versions of Scheme. (I don't know if anyone actually needs first class support for continuations. Does any other language even offer it? However, lexical closures, available in a bunch of languages including SmallTalk, are pretty nice. SmallTalk's use of them to define control structures is perhaps a bit extreme, but not without precedent and I guess is consistent.)

So what provokes the "eh!" reaction (or non-reaction) as it were? (I suppose it was a bigger deal when it was one of the few object-oriented languages, not to mention one of the first rapid prototyping environments.)

The other way of looking at it is that your manager could be a Forth enthusiast. I like Forth a lot too and I have done some non-trivial work in Forth. However, the evolution of computer languages have, for the most part, taken a completely different path. (Again, though, you have words baked into the language for defining your own control structures. Also, one can interpret the ANSI standard to give you the moral equivalent of lexical closures. All of this makes me very happy.)

The language I program in for the fun of it though is currently OCaml.