QUOTE : The thing about Web companies is there's always something severely fucked-up. There is always an outage, always lost data, always compromised customer information, always a server going offline. You work with these clugey internal tools and patch together work-arounds to compensate for the half-assed, rushed development, and after a while the fucked-upness of the whole enterprise becomes the status quo. VPs insecure that they're not as in touch as they need to be with conditions on the ground insert themselves into projects midstream and you get serious scope creep. You present to the world this image that you're a buttoned-down tech company with everything in its right place but once you're on the other side of the firewall it looks like triage time in an emergency room, 24/7. Systems break down, laptops go into the blue screen of death, developers miskey a line of code, error messages appear that mean absolutely nothing. The instantaneousness with which you can fix stuff creates a culture that works by the seat of its pants. I swear the whole Web was built by virtue of developers fixing one mistake after another, constantly forced to compensate for the bugginess of their code.