My last blog entry, REAL programmers don’t use pencils sucked slugs. While many of you may already agree with this statement, there are a few people out there who may feel differently. Somehow the article managed to garner 50 more positive than negative votes on reddit and at least one person blogged favorably about it (bless you Mike Griffiths). Still, it never felt right from the moment I wrote it. I went back and reread it shortly after posting it (I need to let a post settle a bit and then proofread it, otherwise my eyes seem to miss obvious errors) and the first thing I felt was: “That’s not what I meant to say!” I made an effort to tone it down by rewriting the last paragraph, and it felt better so I left it. Still, it never seemed like it was saying what I wanted it to. After about fifty or sixty scathing comments at various locations it became clear why: I set out to write an article about how to be a better programmer and ended up writing one that said that unless you could do complex tasks entirely in your head you probably sucked.

I’m not terribly concerned about the large numbers of negative comments. When someone says something like: “This guy is an idiot, I bet he couldn’t write a linked list.” I couldn’t care less. Anything that creates the slightest bit of controversy and is well read will get these kinds of personal attacks. I’ve gotten strong negative reactions from articles like The Temple of Java and Freedom languages. I loved both of those articles (and yes, the first one was just a lighthearted rant, that’s why it was posted under rants). The difference is that in both cases along with scathing comments I also got some very interesting and useful conversation that indicated people had understood my point. I got a lot out of the comments and cross-posts. I can’t really say that is the case with the “REAL programmers” article. The few positive responses were luke-warm and often related to a sub-point rather than addressing the central thesis. As a writer the definition of bad writing is when people miss the point you are trying to make. This isn’t their fault, it’s yours.

After giving it some more time and rereading it in a more reflective spirit, here’s why I think I managed to lead people astray:

1. The title. I picked the title, “REAL programmer don’t use pencils” as an attempt at some bumper-sticker humor, like “REAL men don’t wear plaid” or “REAL women drive Humvees.” It wasn’t meant to be taken seriously, just to attract some attention. Of course, side-titles like these need to be either tangentially true or explained away within the article. For example, in Making a Ruby benchmark 51x faster than PHP The article isn’t really about Ruby being faster than PHP, so one of the first things I say (very clearly) is that I don’t think Ruby is faster than PHP. That was also a weak title, but I handled it better. A title like Ego Mountain is safe, because it doesn’t make any real statement about what the article is about.
2. I never state my position clearly until the conclusion. It’s fine to lead in with an analogy, but at some early point you need to clearly state your position. Failing that, people will naturally be forced to draw their own conclusions about the point you are driving towards (or simply take the title at its word).
3. I use a controversial position (that learning to do math in your head makes you better at math) as an analogy for another controversial position (that learning to envision the behavior of code in your head will make you a better programmer). Worse, I overstate the first point by implying that you CAN’T be good at math unless you can do a lot of math in your head.
4. I got sidetracked by semi-related and ALSO controversial points. Both the discussion on comments and my postscript about Fizzbuzz did nothing to help make my point. They just piled controversy on controversy.

None of this changes my belief that learning to program in your head is a good exercise to help you become a better coder. Nor does it change my belief that learning to do reasonably complex math in your head will help you to get better at math. Neither of these equates to a belief that once you develop these skills you should stop writing things down altogether, just that they will improve your speed and your general “feeling” for the skill. I’ll probably go back and justify these positions in another blog at some future point, but for now I’m going to leave that alone.

Interestingly, I actually like quite a bit of the specific writing, I just don’t think it makes my point very clearly. I supposed this is analogous to writing good code that later turns out to be useless to the user. The pieces all work, but the whole is less than the parts.