I've been wondering about that lately. Linux is, in theory, a much more secure OS than windows. But how many Linux users consist of people like me that spend days at a time running commands and scripts they find down the back of the internet trying to make bits of software work without any real clue of wtf they are actually doing? 

Depending on your level of knowledge, running a command is less of an issue because you can see what you're inputting (and hopefully have some idea what it does, but often not), running scripts though is a different issue since I doubt most people check every line.

Could have a format c: in there somewhere.

It does stuff. With dollar signs. I get dollars from it. Your fragile ego gets a wee boost. Win, win, win.

I generally avoid doing stuff like that on mission-critical (e.g. L4D) systems. Usually if there's much doubt or risk I'll test it out in a VM first.

Installing node.js (I think) the other day on a Pi and I curl'd a script to do all the installing. Didn't even look at the script first. Figured it'd be OK, cos after all I haven't even changed the default user/pass on the pi yet  :-O~~~  :-O~~~  :-O~~~  T_T