- Feature Articles
- CodeSOD
- Error'd
-
Forums
-
Other Articles
- Random Article
- Alex's Soapbox
- Announcements
- Best of…
- Best of Email
- Best of the Sidebar
- Bring Your Own Code
- Coded Smorgasbord
- Mandatory Fun Day
- Off Topic
- Representative Line
- News Roundup
- Editor's Soapbox
- Software on the Rocks
- Souvenir Potpourri
- Sponsor Post
- Tales from the Interview
- The Daily WTF: Live
- Virtudyne
Admin
Agreed. Which is why stuff like this can only be a small part of an interview process. Give me 100 applicants and I might be able to come up with 10 worth bringing in for an interview. Of those 10, maybe half will warrant a skills test.
If the entire interview process is stuff like this, I wouldn't want to work there either...
Admin
A variation of it. Not quite that obtuse though, I don't think...
I ask a lot of questions of those who work for me though. Not to question them, just to ascertain why they did something. In those cases, the only unacceptable answer is something like, "Because that is the way I have always done it."
I have also been known to ask questions where the only real answer is, "Because I am not fucking stupid", and that is the answer I expect. :smile:
Admin
I've also been told that it's called software "development" because it's never done (think version 2.0, 3.0, etc.).
Admin
I am glad you didnt...
Admin
You may or may not be surprised how often custom software bids fall apart when you discuss monthly maintenance costs. People are surprised that there may be ongoing changes needed, as if we can know just exactly what edge cases will come up, or changes to standards, or just how big of an idiot they will hire and turn loose on the software you build.
Software is all about iterations...
Admin
Oh, I'm well aware that good software costs good money, and that the tar paper shack builders have ruined that expectation.
Admin
my work here has already been done.
also giant space penises
Admin
Are you one of that crowd?
Comments should give you the big picture of what the code does, not repeat it in pseudocode.
Well, sure, but you do get at least some numbers. You can make an educated guess, not a random one. A project that is a little harder than the one which took a week might take three days, might take three weeks, but generally won't take you seven years and a group of researchers.
The problem is, with Fermi estimation, you're pretty much verboten from any kind of educated guessing. You need to pull the data out of your ass.
Admin
I had a CTO who hated comments so much we had to pull them out of generated boilerplate (like laravel migrations) and out of included libraries
Admin
I'd hope that at least the number of continents and the land to sea ratio could be considered common knowledge. "Half the continent" could easily be a wild guess (I have no idea how accurate it is, but it has to be some ratio between 0 and 1, after all). The only other guess in there was the figure for "rainfall after evaporation". If you want to consider that suspiciously accurate, feel free, but you can't realistically claim "many of these numbers" to be from research.
Admin
I hope the expected answer was "I do not know." Hiring people who recognise they do not know everything is a big plus, compared to people who agree to deliver something and fail lately and off budget.
Remind me of the question "What is the weight of a Boing 747 ?". A pilot of such plane would ask "empty or fuelled, what kind of transported good... ". (they have to know it, and it's in the book). But for a web-developper, the answer is "I do not know." (and not any wild estimate)
Admin
There was an example question I saw in a interview tips book. How many sheep would you need to have on an island to sustain a dinosaur indefinitely? The 'right' answer is all about sheep breeding and so forth, but importantly it doesn't say anything about what kind of dinosaur. Obviously, that matters - a giganotosaurus is going to eat a lot more than a coelophysis. So my answer, if asked that one, would start by pointing that out and then assume it's a diplodocus. Of course I'd then go on to describe an interesting solution with an allosaurus or something, but as a mathematician, I'd bring shame on the science if I failed to exploit an opportunity to make a choice that reduces the problem to a trivial one.
Admin
I knew several CEOs who did that. Many people have discovered that finding a way to say "you're awesome" without too obviously laying it on with a trowel, is a good way to advancement.
When I was doing experimental psychology at U, one of our lecturers taught us recruiting techniques. One of his key points was that many, many interviewers grossly overestimate their own ability to interview effectively. Standardised tests are far more effective. Unfortunately many recruiters look at recruitment as a kind of self validation ("my judgement is really important to the success of the company") that they are constantly thinking up new interview questions that are not properly validated.
The worst interview I ever had was many years ago when I was still a hardware/firmware designer. The interviewer, a physicist, presented me with a simple discrete circuit and said "what does that do?". I had no idea. "It's a voltage regulator" he said. "Oh no it isn't," I said thinking this was a test, "There is no temperature compensation on the reference junction, and the amplifier gain is so low the output will be very current sensitive. Also the amplifier has no feedback so the voltage output will depend on the gain of the output transistor." There was a silence. Then he said "I designed that circuit" and the interview was over.
Admin
See J.P. Lewis' paper on large limits to software estimation -- complexity theory has a thing or two to say to those Project-infatuated middle managers...
No kidding -- [the place I interned at](www.bhmi.com) still works with [their very first major customer](www.greyhound.com) over 20 years later, and they're [**proud of it**](http://www.bhmi.com/greyhound.html). Custom software development is all about cultivating that long-term team relationship with your customer; however, in practice, that gets thrown under the bus by short-sighted, tie-strangled sheeple whose eyes are clouded by dollar signs. You also have to recognize *how to find what you don't know*. My answer is "I can give you a first-order guess based on what I recall offhand about 747s; if you want something more refined, I can check the Airport Planning document or the FCOM Limitations section, given more details about what flavor of 747 you have on hand." I agree here -- while interviewing still has a role to play to weed out gross cultural mismatches (Nerf fights would never take hold where I work), a standardized test is a far better tool for getting rid of the incompetents than even FizzBuzz is.Admin
You are quite correct, I have no idea why I bolded that first part. :laughing:
But my statement remains very true on the other statement. What's to say it's not 1 cm, or 100 meters?
Admin
That kind of practical software engineering, yes. I used to be able to do it quite a lot and get reasonably accurate. But that's because it is a familiar problem set. Asking me about the Amazon would elicit the response of "This isn't appropriate for a Fermi estimation because all the variables will be guesswork, so the correct answer is to look it up using Google." Ask me about the sizing of a server needed for a messaging service, tell me the domain (Fortune 500 companies, social media, SME) and I'll come up with a function, suggest a range for the variables, and estimate an error band.
The Amazon question, if the guesses as to the reason are correct, is no good for looking at approaches because it doesn't test whether the interviewee has concrete knowledge and experience of the company's application area, experience which may take years to acquire. Perhaps I'm very good at quickly putting together a very woolly model and bullshitting the inputs, but that doesn't prove that, faced with the need to produce working code, I can use external inputs, experience and fast research to make something that will actually work and scale.
Admin
My one time research director resigned and went back to academia when the bean counters told him that they only wanted a feasibility study carried out if the project had at least an 80% chance of success.
Admin
Yeah, but domain-specific knowledge is an order of magnitude or so easier to teach a new hire than how to think about problems. Someone who is great at breaking down problems can always be taught your specific domains, and will only get better with time. Someone who doesn't know how to think for themselves will always be dependant on your hand-feeding them answers.
Admin
And will follow you around like a lost puppy.
Admin
At my last company, the bosses had no idea how software development works. I can't count how many times I heard "Oh no, we can't start work on version X.(Y + 1), we haven't hit a ROI on version X.Y yet!" I suspect one of the reasons my team was canned was they finally realized the implications of the "maintenance" phase of the software lifecycle.
Admin
I worked with someone like that who has a stutter.
Every five minutes: "I have a qu-qu-question..."
Admin
That is a lot of stupid packed in to a short statement.
Admin
Not going to lie, I've always had a huge problem with this line of thinking. I've literally lost sleep over your insights on hiring for the past few days. Then something clicked.
When I took that fucky job, I knowingly turned down the offer with better fit for the one better pay. And it blew up in my face. I emailed that other (better fit) company this morning on a whim, basically said "At that time, I honestly thought you were the better fit, said that the last one didn't work out, and wondering if by chance they needed someone"
I'll be damned, they did, and needed my exact skillset to boot. Within the hour I got an interview, and was hired less than 12 hours later. Not only that, this time around I got better pay than the fucky job.
Thank you.
Admin
True shit? I am just glad I could help man, but you did it all. You heard something, evaluated it and applied it to yourself. That is all you.
Money means fuckall without happiness. The only time you should take a shitty job that has higher pay is when it is a means to an end and will result in more happiness long-term
Admin
The thread where I was talking about buying a house in a crazy market is :twisted_rightwards_arrows: that way :laughing:
Admin
FTFY
Admin
First consider the original amount of rain that falls without considering evaporation — anything less than about 10 inches (30 cm) of precipitation per year is a desert, or nearly so; we can definitely rule out 1 cm. The highest rainfall on earth is on Kauai, Hawaii; I don't remember the exact record measurement, and for the sake of this argument I have not researched it, but it's 400-something inches — 10 m is a close-enough estimate. Thus, we can also eliminate 100 m from the estimate.
Based on this, I think 1 m is probably on the low side as an estimate for the Amazon basin. Somewhere between 1 and 10 is probably better — since 10 is an extreme, I'll split the difference and then fudge toward the low side; call it 4 m.
Now for evaporation, absorption into groundwater, etc. Rain forests are humid, and in the tropics, hot. The heat increases evaporation; the humidity suppresses it (but the humidity came from evaporation in the first place, so there must be quite a bit of it). Here's where the
guessingestimation gets less accurate. To me, 50% evaporation/50% run-off seems like a reasonable guess, which would give 2 m worth of rain flowing into the river, but it could easily be 75%/25%, which would match @ProbablySignedUp's estimate of 1 m, or 25%/75%. In any case, 1 m is almost certainly within an order of magnitude of the correct answer.Admin
Well then, what is the airspeed velocity of an unladen swallow?
Admin
Do you have that memorized, or did you look it up?
Admin
Having now done a bit of research, it turns out the Amazon drains about 40% of the South American continent, so @ProbablySignedUp's estimate was a bit high, but not much.
I didn't find a rainfall figure for the Amazon basin as a whole, but based on the following numbers, it's probably a bit over 2 m, average: Manaus, Brazil, in the heart of the Amazon basin, 2.3 mr; Rio Branco, 1.9 m; Iquitos, in eastern Peru, 2.9 m; Santa Cruz, Bolivia, 0.9 m. Still no clue how much of this winds up in the river, so 1 m net remains as good a guess as any.
Admin
African or European?
Admin
Huh? I don't know that.
Admin
Yup.
Agreed. Lesson learned.
Admin
I didn't look anything up until after I posted that. I know how much rainfall there is in a desert; I grew up in the arid climate of SoCal. I remember the raininess of Kauai for two reasons. One is an atlas we had when I was a kid; it had a page of facts like highest and lowest elevations, maximum and minimum temperatures, maximum and minimum rainfall, etc., for the world and each continent. The other is that I have been to Kauai and took a sightseeing helicopter ride around the island, which included flying up to near the summit of Mount Waiʻaleʻale. I admit to having to look up Mount Waiʻaleʻale for this post, as I misremembered the name.
Admin
Then you have an advantage when it comes to useless interview questions. :laughing:
Admin
I am happy for you man. Way to go on landing the job. I hope it works out well for you. I hope you are going out to celebrate tonight?
Admin
Now I'm wondering about the decision that I made two years ago.
I had two potential jobs.
The first was less money, a dedicated programming position, and most importantly, a marketing company that couldn't exactly explain what they do when asked.
The second was more money, more of a help desk position, and a company that provides healthcare services, in a specific field that I can get behind (Belgium cancer).
I took the second job, but I think that I might be personally less happy because of what I do. :\
Admin
If a company cannot explain, in one sentence, what they do...you should shy away. Are they still around?
Admin
If they were my client, I would definitely not want to piss them off as they'd surely throw me under the bus.
Admin
Some of my former coworkers (from when I worked downtown) have a hard time understanding why I stay at a job in the suburbs where I can roll in at 10:30am dressed casual. They also can't get why I wouldn't want to pay twice as much rent as I'm paying right now.
Admin
All young, I assume? That seems to have a high likelihood of having "omg cities are cool!"
I live in a boring neighborhood 1.4 miles or so by car from work. Both places are a couple of miles outside the first Dallas ring road. What I give up in not living and/or working downtown, in addition to the things you mention, are, among other things, nice second-order effects like "7-minute commute" and "I get to go home to eat lunch every day if I want."
Admin
Mostly late twenties. Oddly enough, when I was that age a few years ago, I didn't want to live in the city. The twin annoyances of higher living costs and hipsters form a very compelling counterargument.
Admin
Close enough.
I did, but I couldn't afford to. But I did spend a weekend at a friend's in Manhattan, who was paying I think $600 or more to sublet the tiny bedroom of a tiny one-bedroom apartment on the 3rd or 4th floor. I don't know how people live that way.
Admin
I do a couple of interviews per week for my company. A lot of our candidates end up fiting the bullet pointed tech skills perfectly, but its hard finding someone that can think even slightly outside of their realm (in other words, innovate or ).
Whenever i see these stories and see all of these people who dont get the point of these questions (despite their poor execution), i wonder if some of you fit. Often the answer itself is irrelevant.
Admin
You guys are all wrong. The amazon question isn't the trwtf here: the trwtf is clearly—and this is even stated in the article—the fact that initech's ridiculously arduous recruitment process has yet to pass a single candidate; not even a secretary. The trwtf is how much easier it is to focus on a stupid simple math problem than to identify and solve people problems.
http://what.thedailywtf.com/t/omg-be-smarter-than-blakeyrat-at-interviews-the-dumbest-blakeyrat-there-is/5144/65
Admin
The problem is the way the question was asked.
If the question was "ESTIMATE how much water flows.....", then it's a good question. But, the question was "How much water flows..."
If the interviewee is wise, they'll be able to answer the first question, but refuse to answer the second. I will not tell you how much water flows out of the Amazon, because I do not know, and I don't know what you'll use the answer for - I don't want to be sued if I'm wrong. I'll make an educated estimate for you if you want though, I don't have a problem with that, as long as you know it's an estimate.
Using 'estimate' in the question makes this clear. Failing to use that word is either meaning to trick the interviewee, or is asking the impossible. In the real world, if someone asked me that question, I would refuse to answer like that, and, for all I know, the interviewer wants me to refuse in the interview as well, for the same reason. Using 'estimate' makes it much clearer.
Interviewers need to stop trying to trick interviewees, and stop trying to hide their inferiority complexes by attempting to humiliate people.
Admin
Why not make it clear what you are wanting? There are ways to ask questions which will help you get the answers you want.
If you want someone to say how to work out the height of a building using just a foghorn, then tell them straight out you just want to know how they'd think it through, rather than saying 'you have only a foghorn, tell me how high the Empire State Building is'. Instead say 'assuming you have only a foghorn, can you talk through how you might work out the height of a building?' I want to see how you think.'
Why do interviewers try to trick/humiliate interviewees? Are they feeling threatened?
Admin
I get your point, but your pedantic answer makes me wonder just how carefully I would have to phrase things if you were in my employ...
This article is a dramatization of an event. I am sure some artistic license was used. They probably should have used more, because to me this is not a WTF. You do not know how it was originally phrased...
Let me give you an example from what we do and what I have asked of my people:
We have been working on a recent project that I want to roll out to the public pretty soon. We are in beta testing right now with some of our other customers. When we started this project, I asked one of my people, "How many users can we support per server?" I did not say "estimate", I just asked the question. One of my guys got back to me pretty quickly with, "I would estimate between 10K and 50K concurrent users per server." Now, that is a pretty wide spread, but it got me the information that I needed.
It sounds like, from your response, that you would have just thrown up your hands and exclaimed, "How am I supposed to know? We have not even started building the damned thing yet! I have no information!!!111!!" Yeah, I don't care. Give me a ballpark. Make some motion. Be brave enough to be wrong.
Now, granted, this was a domain-specific question. But there was next to no information available to him.
If I asked this same exact question of a person in an interview, "How much water flows through the Amazon river every year?", and they just throw up their hands and give up, we move on to the next applicant. If they respond with, "Hell if I know, but here is how I would go about finding out..." and provide a coherent answer (that does not involve Googling for the exact answer), they would progress.
I feel it safe to assume that they are not going to be asking for mission-critical answers in the preliminary interview process and that liability would be pretty fucking limited before we even hire someone. But hell, that is just me...
Admin
hmm... good point.
in that case the correct answer is provably: "All of it"
The seais a collective, multi word, noun for all of earths interconnected oceans, gulfs, seas, bays, basically all interconnected saltwater areas.The Amazon River has a number of mouths, but all of them drain into
the seatherefore all of the water in the amazon river flows into the sea.
not a terribly helpful answer, but it is a correct answer!
Admin
The problem is that this is a question in an interview. If you are not used to this type of question (note NOT this type of thinking), then is is not unreasonable to assume that a 'tell me this' question in an interview expects a correct answer, not a guess.
Are you trying to recruit someone who knows how to do the job and can think, or are you trying to recruit someone who knows how to answer interview questions?
Interviewers need to stop trying to trick candidates. Just because you, as the interviewer, know what you expect, you can't assume that an interviewee (who is already stressed and feeling under pressure) can read your mind. Just tell the candidate what you want from them, don't try and trick them to make them feel belittled.
What do you lose by explaining what you're trying to achieve, other than causing confusion to the candidate?