среда, 27 апреля 2011 г.

как троллить на собеседованиях

воображаемое собеседование злобного тролля Ричарда Фейнмана в microsoft (со всеми их дебильными задачками про 3 лампочки):
Interviewer, exasperated: You could turn two switches on and one off. Then wait a minute, and turn the third switch off. When you go to the other room, the lamp controlled by the first switch will be on, the lamp controlled by the second switch will be off and cold, and the lamp controlled by the third switch will be off and hot. That seems a lot easier than all this rigamarole about disassembling the switches or building custom equipment.
RPF: How am I to measure the heat of the lamps without special equipment? You just said that I couldn't reach them.
Interviewer: Um. Yes, I suppose I did say that.
RPF: I can see a number of additional problems with your heuristic. You haven't specified how far it is between the rooms, but have several times implied that it is a considerable distance; I can't see the light from the switch room, I can't align my mirrors and I can't bring a conductor that long for continuity testing all imply considerable distance between the switch room and the light room. The time it takes me to get from one room to another can give the third lamp time to cool. The third lamp might not be very hot to begin with; if the fixtures are fluorescent bulbs, as they are in this building, or modern LED bulbs, then their heat output is far lower than an incandescent bulb. We also haven't specified where this scenario takes place. If it is in a very hot climate, like Los Alamos in the summer, both de-energized lamps could reasonably be warm to the touch, and if it is in Alaska in the winter in an uninsulated room then both could reasonably be cool by the time I get there. Your proposed heuristic depends upon a number of conditions that were not given in the problem. And it is in general a bad idea to test whether something is extremely hot by touching it.
Interviewer: Well I think that concludes this portion of the interview. Before we let you go for the day do you have any questions for me about this company, this team or the job?
RPF: Yes. When you build software algorithms, do you build systems using well-established software engineering principles to produce software that conforms to industry standards and practices?
Interviewer: Of course.
RPF: And do you use software analysis tools, like profilers, debuggers, theorem provers, and so on, to facilitate detection and diagnosis of flaws?
Interviewer: Yes, again, of course we do.
RPF: Then why would you ask an interview question that tests my willingness to abandon industry-standard, well-established techniques that use common electrician's tools to determine continuity of a portion of an electrical system? And why is the solution you were clearly driving me towards one which takes advantage of an undocumented and unreliable epiphenomenon? Does your team usually write code whose correctness relies upon undocumented and unreliable correlations, correlations whose magnitudes can vary widely as a result of implementation details?
 ржал как подорванный например

1 комментарий:

  1. Смеялся от души.
    Версия на русском: http://blogs.msdn.com/b/ruericlippert/archive/2011/02/28/what-would-feynman-do.aspx