Glenn Vanderburg

: Blog

infosyndicate
1 of 1 article
Try It
Sat, 12 May 2007 (19:59) #
Two of my best friends in the Ruby community have suddenly gone all nyah nyah on us, and it’s time for a bit of reality. Neal Ford says static typing is communist bureaucracy, and Stuart Halloway (presumably trying to tone things down a bit) says it’s more like training wheels.

Being intentionally provocative to knock people out of old assumptions is all well and good … but even though I love Ruby, and am so glad to be working in it now after 11 years of Java, my perspective is more pragmatic.

There is no perfect world. You gain a lot by giving up static typing for dynamic typing, but you also give up some things. It’s six of one, a half-dozen of the other. Or maybe 9 and 3. Whatever. Whether the tradeoff works in your favor depends on a lot of things — the kinds of systems you’re building, the process and practices you use, how well you understand your domain, and other things. And folks like Tom Ball make great points when they point out some of those tradeoffs.

(Furthermore, "static typing" doesn’t necessarily mean "Java-style static typing." It’s just that the industry right now is all wrapped up in static typing done the wrong way, while ignoring static typing done the right way.)

In spite of all that, though, I do agree (mostly) with Neal and Stu. I, too, would rather write Ruby in vi (er … that is, emacs) than Java in IDEA.

If you’re still on the fence about this, or even downright skeptical, here’s my advice. Having been a static typing bigot for a very long time, I know all of the arguments from that side. I used them myself. And what I’ve learned is that you can’t figure this out by thinking about it. There are things about dynamic programming languages — and about how programmers work in those languages, and how systems are designed in those languages — that you don’t really understand until you’ve gained some real experience. It’s the Blub paradox at work. Knowing about the flaws in the language isn’t enough either. What’s important is knowing when they bite you, and how badly, and what you have to do to stay clear of them. Primitive types in Java represent one of that language’s worst flaws, but in my many years of Java programming I was impressed by how rarely they really got in my way. (They did get in my way, but not as much as I expected them to when I first learned Java.)

So just give Ruby (or Groovy, or Python) a try. Find a way to build a system of some reasonable size (a few thousand lines or so) in it. Give yourself time to get over the learning curve of syntax and names, and learn to think in the language. Decide based on experience, not fear or doubt.

Our field is at its worst when we act (when we teach, design, manage, plan, argue) as if we’ve got it all figured out, as if we’ve got it down to a science. Nothing could be further from the truth. Our field is one of creativity, experimentation, guesswork, hunches, and feedback. Above all, feedback.

Nothing blinds us and deafens us to feedback like being convinced we’ve found the one true way.