|Ned Batchelder : Blog | Code | Text | Site|
Languages and IDEs, again: Just write some code
» Home : Blog : November 2004
I read Simon's complaint about Eclipse this morning, and spent a little time looking into it. I considered downloading the latest milestone, and I investigated what version I was running, and whether it would upgrade itself, and so on. And then I realized that fiddling with Eclipse was not how I wanted to spend my software engineering energy. I wanted to write code. And I thought about The IDE Divide, which says you're either a language person or an IDE person. Bob responded to that division of the world by saying, "you need both". I guess my attitude toward the dichotomy is that you can do really well with neither.
I often find myself tempted by the glittering riches of some new IDE, or some new language (or language feature), when what I really need to do is buckle down and write some code, or even harder, do some of the difficult thinking that happens before coding. Usually, when I am not making progress on a project, it is because of something that neither languages nor tools will help. Usually, I am mulling over some thorny problem, either in the user experience, or in the architecture. When those hard chunks are out of the way, it's pretty clear sailing.
So I didn't do anything about Eclipse this morning. I continued to hack Python in one of the many Python IDEs that clutter my hard drive. Aha! "Many IDEs" must mean I'm an IDE guy. I'll admit it, I like playing with new tools. But the fact is that I use all those IDEs the same way: as a Python-aware text editor with tabs. I don't do much more with them than that. For me, IDEs occupy a similar slot to candy: they're tempting, almost irresistible, enjoyable for a short while, but ultimately unsatisfying. They don't get at the heart of the problem.
I haven't had an opportunity to use Eclipse's luxuriant refactoring tools and quick fix doodads. I'm sure they make developers more productive, how could they not? But they won't help me decide how to refactor, or what the right semantics are for an abstraction, or predict in which ways the system will have to change in the next release. They won't help me find the simple path among the complex, or choose just the right words to describe my thoughts, or understand the user's problem better. They may help me be a more productive coder, but they won't help me write better software.
So, I must be a language guy, right? Well, no. I haven't dug into Ruby or Haskell or Smalltalk, and frankly, I don't know why I would. I can't explain how Scheme is different than Lisp, I have only a primitive understanding of what a closure is. Even in the languages I use, I am slow to adopt new features. I haven't installed Python 2.4 yet, and I haven't incorporated 2.2's generators into my daily routines.
I'm technologically skeptical and slow to adopt in general (shameful confession: I don't have a TiVo, an iPod, or a cell phone camera). I'm interested in solving problems, and if the tools I already have can help me solve them, I'll stick with them.
The fact is, there's no shortage of choices out there, whether it's IDEs, or languages. For that matter, what of the support facilities we use during development? Just in the Python world, there are many unit testing frameworks (at least four), many documentation generators (at least ten), many parser generators (at least 19), and so on. Using any of these requires investigation and learning, and correctly choosing among them requires investigation and learning multiplied many times. My time is precious, and sometimes I just need to cut off all the exploration and ask myself what I've actually accomplished in the last hour. When I'm stuck on a particular difficult challenge, those twinkly tools are an especially dangerous distraction. Experimenting with a new toy is easier than facing down a hard choice, and playing with the tool almost certainly will not provide the answer.
Of course, I need to stay abreast of the possibilities. Most of these things offer real benefits. A parser generator is a huge win over trying to write a parser by hand. But all that glitters is not gold. Sometimes I need to stay focused on what really matters and put my nose to the grindstone and write some frickin' code.