1 Dr No: Take him to the cell!
1 Henchman: Yes boss! Er, what cell?
2 Dr No: The escape-proof cell!
2 Henchman: The electrified one?
3 Dr No: Yes the electrified one! And don't give it away! {The henchmen escort Stud away.}
4 Dr No: {shaking head as he mutters to himself} Henchmen. Evil, obedient, bright - pick any two.

This refers to the common adage in software development (at least that's where I heard it - it apparently is common in other fields of research and development too): Fast, cheap, good - pick any two.

The idea is that you can't get everything you want when developing something:

• If you want something done fast and cheaply, it won't be any good.
• If you want something done fast and well, it will be expensive.
• If you want something done well and cheaply, it will take a long time.

It took some searching to find this in Wikipedia, but I eventually found it.

---

2016-12-21 Rerun commentary: Apparently this principle is best known in the field of project management in general. I was just pondering whether anyone has thought of expanding the triangle to some sort of tetrahedron, and lo, of course someone has. Not just one person either. The fourth node is apparently "scope", which seems to be roughly how fit for purpose the end product is. So if it's relatively fast and cheap and good quality, it's not actually fit for the job you wanted it to do (it might do something different, or in a way that requires you to change your workflow, for example). If you want it to fit exactly the specifications you set out, then you have to sacrifice one or more of development speed, expense, or end quality. (At least that's my understanding from 2 minutes of reading those pages.)

Eh... as much as I'm in favour of needless expansion of two-dimensional concepts to three dimensions, it kind of loses some of the elegance of the simple triangle.