My Startup Job Interview
blog.galler.io
blog.galler.io
Great description of surprise push-to-production.
It turned out that the department had carefully orchestrated a system whereby all checked-in code would move automatically to alpha the next day, then beta, then production the next week. But nobody told me.
A manager there later told me, "If the code we write is strong enough, it does not need to be released--it will escape."
And the student understood.
- That the manager decides what gets pushed into production?
- That the manager decides not to inform the programmers concerned?
- That you have a clown for a manager?
- That it's probably time to quit the company?
"You have 60 minutes to code a security patch for a vulnerable payment processing app."
"Congratulations, you shipped to a million people and saved the company from hackers."
I laughed pretty hard at that, and the "aggressive integration bit"...I'd probably leave if a company decided to ship something I coded in an hour in an interview.
I know there's a bias here (better engineers are probably more likely to blog in the first place and most people would rather write about an experience where they're the "victor"), but it would be nice to read an article from someone acknowledging their own failure just once.
There were loads of interviewee issues that occurred in the article. Hopefully that balances things a little.
edit: HN discussion: https://news.ycombinator.com/item?id=6243627
There's a good chance the next company will suck for some reason or another. On to the next one.
Maybe it will help everybody understand that this is a satire...
Sometimes it's best just to laugh at the sheer insanity of it all. We'll get there guys, but it might take a couple hundred years.
In light of this being satire, the HN news comments are 10x funnier than the article itself.
Some unsolicited advice (you're lucky, you found out it was a crap company in 2 hours) and some PC outrage (why does it have to be an Asian lady).
I'd also never seen software deployment by RAID before... (or happily since)
Mind you, they were bought for a pile of money and the people who thought the above were a good idea made out like bandits.
I read it first as a description, not satire.
The number of rabbits at the (i+1)'th step is equal to the number of rabbits at the i'th step minus that one rabbit which is /maybe/ being eaten. And the "maybe" I think can be quantified as the probability the Troll has to find a rabbit at the i'th step. So:
R_(i+1) = R_i -1 * P(Troll finds a rabbit)
And now we have to calculate the probability of the event "Troll finds a rabbit" (i'll call it event A). There are 2 cases: * the Troll falls into a black hole (btw, i'm gonna assume the blackholes teleport to any cell with uniform probability); * the Troll does not fall into a black hole;
I'm going to call these two events B and not(B).
So:
P(A) = P(A given B) + P(A given not(B))
Well, P(A given B) is the probability that Troll is being teleported to a cell with a rabbit in it, that is number of rabbits over number of cells: R_i / (nm)
P(A given not(B)) is the probability that Troll will step on a rabbit in a nearby cell which is: rabbits' density over 8: R_i / (8nm)
So:
R_(i+1) = R_i - ( R_i/nm + R_i/8nm ) = R_i * ( 1 - 9/8nm ) = ... iterating ... = R_0 * ( 1 - 9/8nm )^i
R_0 being the initial number of rabbits. I'm not sure all this is correct, but that formula confirms the following intuitions: * The number of rabbits is monotonically decreasing at every step; * The bigger the grid, the slower the decrease.
edit: The closed formula would be:
int numOfRabbits(int numOfSteps) {
if (numOfSteps == 0) return R_0;
else return R_0 * (1 - 9/8nm)^(numOfSteps-1);
}It's a silly question unless it's for a quantitative role.
A closed form expression is basically just a mathematical function used to model a potentially infinite series.
Take a look here: http://en.wikipedia.org/wiki/Closed-form_expression
Look at the first and second examples.
You'll see the big summation sign (for loop: i to infinite) and then an explicit function.
If you want to think, what the heck is this math crap used for: think about a more basic everyday concept like recursion. How do you know it will terminate?
There was no information whatsoever about what changes on each iteration... if nothing changes then the answer would be a constant of whatever rabbits were at the beginning... and that doesn't sound about right :/
As a first go I'd assume that the topology was that of a torus, so as to not deal with the boundary. Now, as the exact definition of a "round" wasn't specified, I'll just assume the troll can walk forever.
Now, it's been proven that a simple symmetric random walk on a 2D lattice has probability 1 of reaching any given point as the number of steps goes to infinity. So, ignoring the teleporting, the closed form answer for the number of rabbits eaten is mnR.
With teleporting, I'd guess the Markov chain would mix better. So, the closed form answer would again be mnR.
E(number_of_rabbits) = rabbits_in_last_turn - prob(land on rabbit) - sum_i_to_infinity(prob(land_on_blackhole)^i*prob(land on rabbit))
But my bet would be zero, if the troll has infinite time
Also, if there is some decent way to search the links or stories which I don't know of, by all means, tell me.
Verbatim means "in exactly the same words as were used originally."
i.e. "This is exactly how I perceive start-up interviews but I've taken some creative liberties to drive my point home."
Just curious.
Why are you saying Dave the Polo Shirt guy was not white? He didn't mention that.. Not saying you're a racist..
Its as if I wrote the following:
John, Mike, and a white man sat down in my table. We all talked about our trip to Disney Land. Dinner was fine, except for the fish. It was over cooked. Mike sent it back to the kitchen, and they prepared another one. Afterwards, we went to the local arcade. Where a mean game of Galaga was played. John, as always, had the high score. I was not as lucky. The score was affected by my bad sleeping habits. Ever since starting this new job, I simply cannot fall asleep as easily as before.
If it had been 'two white guys and a tall woman' I probably would have pictured the interviewee as non white. I also imagine (because its in blog/HN comment format) the objection to singling out the asian. That makes it seem honest, because noticing that she's asian is like an admission of guilt.
This story is making fun of stereotypes. Using clichés is how he's getting a particular effect.
Deleted comment
And who cares what they were wearing?
And why do we care what color the fonts were? He could have just said it was hard to read.
Hahaha.
Really any more description than that and you're being something-ist!
That story sounds fascinating, tell me more!
It was a story. I don't think we want stories in general to be devoid of any ornaments at all.
If you live in Japan (as a white male) you will be commonly referred do as a 外国人 (foreigner) by every Japanese person.
Does that mean that all Japanese are racist? No.
[1]: http://goodmenproject.com/featured-content/why-is-white-male...
>But now we have female doctors, lawyers, professors, and presidents. In fact, we have so many of them (even in heavily male-dominated areas like the physical sciences) that it’s no longer fair, or correct, to assume that the person in question is male. So how about we just collectively get over gender-stereotyping truck drivers, cardiologists, and pre-school teachers? I’m not saying we should aim to stop using gender qualifiers right this second. But it might not be a bad idea to pause and ask yourself why you think you need to specify that the doctor is female and the teacher is male. Is it important? Or is it just not exactly what you would expect?
PC, while good when faced with derogatory remarks and such, can be overused.
I did my bachelors degree in a pretty small school which was mostly filled with white teachers and students (I live in the north), and on staff there was only a single black man. It was quite interesting repeatedly seeing PC conscious people looking for him:
"Is Andy here?" "Which Andy are you referring to? We have three," "Oh.. erm.."
I am not saying "Oh, lets all be non-PC", but if the only defining feature about a person is being asian I dont see why it would be any worse than saying "dark haired woman".
The minute we assign some kind of shame in either direction when it comes to pointing out race, disabilities or any other defining features, that is the moment we are steering in the wrong direction.
Joking aside, it's possible that these details stuck in the OP's mind because she happened to be one of the only females and/or Asians in the group.
Just curious?
Someone in the audience asked if they made the interviewees sign anything since the company was using their unpaid labor and the guy giving the talk was seriously dumbfounded as if the idea had literally never even occurred to him.
I want to be smarter.
"My honesty"
"I don't see honesty as a problem that needs changing"
"I don't give a shit what you think"