[colug-432] the /etc test

[Ethan Dicks]

On Fri, Jul 25, 2014 at 12:52 PM, Tim Randles <tim.randles at gmail.com> wrote:
> I was discussing this thread with a coworker just now and shared with him my
> thoughts on google's interview process.  I don't like it as a general
> example of useful interview practices.

I would tend to agree, generally speaking.

> In my experience I find that asking questions that probe the boundaries of
> the expected level of knowledge for the position is much more useful than
> asking abstract questions.

My "favorite" questions start off with, "tell us about a time
where..." to let me draw on my experience in a variety of situations
that are quite concrete.  Examples include "the entire building and
datacenter lost power", or "a customer called at quarter-to-five as
you were leaving for vacation" (I definitely had a story for that
one), or even "you had to give negative feedback to another member on
your team about the code they'd submitted".

Admittedly, this format works well for more senior positions, but I
think that's part of the Google approach - they interview a *lot* of
fresh-out-of-college and still-in-college-but-graduating-soon folks
who _won't_ have professional experiences in their background, and
are, in fact, used to having to emit code on demand for bizarre and
artificed situations.

> It also allows the interviewee to give their best shot at
> answering and empowers them to admit when they don't know something.

In the interviews I've given over the years, I've never been expected
to know *everything*, and I think it's good to get a glimpse of how
that will present down the road.  When I am interviewing, I jot down
notes not just of what was asked and answered but how.

>  Asking
> an abstract question that may or may not have an actual solution often
> leaves the interviewee struggling to formulate any answer.  It can cause
> their confidence to plummet and that does not help me evaluate their true
> abilities.  I want the interviewee to have full command of their faculties.

This is where I guess I don't quite get it... is that the common
reaction to being asked a question you don't know the answer to?  To
lose confidence?  If you come into the question knowing there isn't a
"right" answer or *one* right answer, I think it leaves you freer to
explore oddball approaches.

> I do have the luxury of full-day interviews for candidates that make it past
> the phone interview.  Personality evaluations are made during lunch and a
> half-hour "coffee with the team" session.

I've only done a few full-day interviews, and I found all of them
(except Google) to be good experiences (there were a couple of
unsmooth spots in my Google interview, including one of the
interviewers having an off day, problems connecting, problems pulling
his questions up, that I think colored the experience.  Got good marks
from everyone else that day).  Lunch is important for considering the
social aspects, coffee-with-the-team is a good idea, if your team does
gather already (some places I've been, that would be an unusual event,
so the team itself wouldn't feel as comfortable as they otherwise
might).

-ethan