So, the difference between a "hacker" and an "engineer" has always been this idea that the engineer is designing something which is principled at all levels, whereas the hacker is putting together something which works in the most obvious cases but doesn't necessarily cover all of their bases. In practice, I think we all know that a little of both is best; ideally you can throw something together with copy-paste programming to sign a new client, say, and then you can rethink it from the ground up once you've got that proof that someone needs it. Hacking is much cheaper than principled design.

What I'm taking away from this article is that this is also true at the CEO level. Conrad sounds like a hacker's philosophy, writ managerial. If there's a regulation that your browser spend 52 hours on a web page, here's a script to make your browser browse that web page for 52 hours, clicking aimlessly among the pages. You're welcome; you've got other crap to do. Or, if you don't have a license to sell this thing -- well, you still seem to be able to sell, so let's just do it the way that works for now, as a proof-of-concept, and we can always do-it-right later.

And that's not a bad thing! It works great for software! But it looks from the article that therefore the hiccup that the company faced was somewhat predictable given the personality of the CEO. It's hopefully an instructive lesson for any venture capitalists who don't already realize that you need that both-hacker-and-engineer versatility up-top. Because it's the exact sort of hiccup you get when your software team is hacker-only and always put under pressure to get more done, but never under pressure to do it right. Eventually the code spaghettifies to deal with all of the real-world problems, and the sooner you can catch it before it becomes a hairball, the better.

My job right now is, I maintain a big data processor. I can freely testify that there are still a ton of quick hacks that are stuck in various places. But every week or so, I get a little downtime -- and when I don't spend that peeking into Hacker News, I spend it trying to modularize the hackery and then swap in carefully-reasoned modules for quick-hack modules.

I implore other software engineers to do the same; and I will definitely start thinking of more management-level decisions this way, in terms of "quick clever hacks" versus "careful design."