All of them are super-wrong.

All of them are long, hard projects. All of them requiere specific skills, that maybe are hard to know because you don't find much info about how do them (for example, I haven't find good enough, simple material in how build a relational engine).

But do it are easy. Because the "science" behind them is more SETTLED. Is just niche.

> YapDatabase, Audiobus, or AudioKit

I don't know them, but it look the same as the things I'm talking about. I LOVED to have the time or funding to devote to this kind of projects and living from them (ie: I want to build a relational language.)

IN CONTRAST

The most "simple" apps, are HARD TO DO.

Them are easy projects, but DO THEM is harder. The specs are unclear, you can't rely in a cool algo that solve most of it, you can't relly in a big, large, solid foundation, you NEED TO BUILD AND PULL from several sources in how do them.

Rocket Science is "solved", but you can waste months trying to finally know what the hell is necessary to build that e-commerce website.

Just look at the madness with JS. Is now easier doing assembler than that.

----

So, I mean that the human factors are the uncertain nature of most software projects are a higher burden that the actual "hard" projects.