Why Are Manhole Covers Round? (on Microsoft brainteaser job interviews) | Hacker News Reader

Why Are Manhole Covers Round? (on Microsoft brainteaser job interviews)

hachettebookgroupusa.com

24 pointsgruseom28 comments

Because they are easier to manufacture than other shapes which will not fall through the hole either: http://en.wikipedia.org/wiki/Reuleaux_triangle

Want to attract nerd tourists ? Build your city with Reuleaux manhole covers in the streets and Penrose tilings on the walkways.

Seen in SF: http://flickr.com/photos/ojbyrne/224171975/ :-)

So you can roll them out of the way more easily. Also - no orientation required. My physics is kind of rusty, but isn't there a strength benefit from the shape as well?

Incidentally, many sewer covers in Australia are square, only the larger ones are round.

Okay, I'm done taking this article literally now. :)

Good god, I'm dumber for having read that.

(Why is the cover the same shape as the hole? Because otherwise it's a pretty shitty cover (i.e. gaps are bad))

It's astounding to me that people still care about these artifacts of another era where people fawned over whichever corporate master they sought to please. Good riddance to that, and if PG had never done anything else, he'd still be able to point to that genius ad with "Larry and Sergey won't respect you in the morning" and remain a hero.

I couldn't agree more. I think it's an asinine practice that, if it is in fact being discredited, deserves it. I posted this as an example of the brainteaser interview being discussed elsewhere (http://news.ycombinator.com/item?id=162672), and am a little disappointed that most of the comments here are actually about manholes.

Brainteaser interviews are an artifact of the fundamental problem: there's no easy way to identify good programmers, so people come up with these ersatz, dysfunctional substitutes instead. The brainteaser is a seductive one because it's easy to think "programmers are smart so let's test for smartness instead". This confuses a necessary condition with a sufficient one. It also begs the question, since smartness isn't easy to test for either. The concept of "smartness" underlying the brainteaser interview is a shallow one of the party-trick variety. I see no reason to believe that it's correlated with the kind of programmer I'd want to work with on a project. Brain-teasing opaqueness is not a good thing in production code, and real programming problems are not much like riddles.

I also agree with your other point. Interviewers probing a programmer for "smartness", like a horse might have its teeth examined, or better, like a cow might have its udders palpitated - just feels unseemly to me. It's part of a corporate culture that is the modern version of the feudal system. It's the way you'd treat a peasant. The genius of the Larry-and-Sergey ad is that it speaks truthfully about this issue of respect, which is a very deep one, and of course is ultimately all about self-respect. (Look at some of the reactions to PG's lion metaphor to see just how true that is.)

So large organizations have no reliable process for hiring good programmers. As Paul Graham's writings have articulated so clearly, the fact that they have that problem doesn't mean that you do; they need you more than you need them. There's actually a much better test for which programmers add value in the long run: the market itself. Pass that test and you won't have to ask anyone to approve of your "smartness".

> "I'm happy to say that we're recommending you for immediate hiring into our marketing department."

It was worth reading this just for the Feynman anecdote at the end.

Complete Feynman interview - http://www.sellsbrothers.com/fun/msiview/#Feynman

Too bad the Feynman story is completely made up. It seemed pretty realistic.

I don't get the bit about the marketing department, though. How does cutting through bullshit constitute an immediate qualification for marketing? The story would make more sense if he were hired immediately for the technical job. But that wouldn't be much of a punch line.

Edit: I was going to say that MS would never have interviewed Feynman, they would have hired him immediately. But actually the anecdote is quite good in this respect. Feynman would have insisted on going through the standard hiring process.

My guess would be that (many, most) manhole covers are round because it is the obvious shape for most lids. There are so many poorly designed urban objects that I tend to believe that the feature of not falling into the hole was a lucky accident.

Think of stop signs, for example. It would be really easy to have a different shape/color for four-way stops, but at least in the US all they do is add "four way" to the sign in an inconsistent way.

I think having multiple shapes for stop signs would be a terrible idea. The red octagon is an international standard. Messing with it is just asking for trouble. The "4-way" signs are just an aid to drivers, to indicate what behavior they can expect from other drivers. They're not required, and no replacement for plain old good judgment.

because manholes are round

I think they are round because circular shape is the only shape that cannot fall into itself.

If manholes are square or any other shape... at certain angles the cover could always fall into the hole

You can make a cover of any shape you want that won't fall in. It depends on how much the cover overlaps the hole. Even with a round cover, you can't have zero overlap or else the cover would fall in.

Now back to the correct answer, which is because the manholes are round - that begs the question of why manholes are round. It's probably a matter of equally distributing the sideways forces from the dirt plus a matter of being able to manufacture round tubes more easily than, say, square tubes or hexagonal tubes.

It won't be a cover unless it's slightly larger than the hole wouldn't it? :)

Actually, no, a square cover would have to have a much wider flange in order to ensure that it could not fall in . (rotated 45 degrees in relation to the whole and tipped up, a square cover would fall in)

Actually, yes, that's exactly what I was getting at. You'd have to make the square cover sqrt(2) times wider than the hole, but you could still prevent it from ever falling in. (Granted, it's not a very efficient solution, though.)

Ok, agreed :-)

Dang it, you got there before I could!

I didn't have a single brain teaser in my msft interviews!

I was almost disappointed because I am usually quite good at them.

When was your interview? In another thread, somebody said that msft had mostly given up on this practice.

2006 I believe.

Round ones are easier to screw on ;-)

Short answer: They're not. At least, they're not round everywhere in the world. There are plenty of square and square-like covers in the UK/Ireland, some of which split in half diagonally. Perhaps a better question is why should they be round or what advantages do round covers have?

If they weren't round, superheros couldn't use them as frisbees/weapons.

If a car hits them and they bounce and spin, they will still (most likely) settle back into the hole correctly. Even a Reuleaux polygon won't always do that.

There are no points more highly stressed than others. Put an extreme amount of weight on them and the force is evenly distributed around the circumference, rather than overly stressing the corners and bending the cover (on a square lid).

Manholes have flanges.

Circular and reuleaux triangles aren't the only shapes that won't fall through.

So why are round shields better?

My guess would be if someone swings a sword at you in a downward fashion it would slide or deflect off a round shield whereas it might just pull you down/knock the shield out of your arm if it hit the top of a square shield.

Corners don't get caught on crap and slashes are deflected tangentially around the circumference rather than cleaving the shield itself.

Thanks. Based on some shield images (http://images.google.com/images?sourceid=mozclient&ie=UT...) it looks like there are non-circular shapes which provide at least some of those benefits as well as possibly improved coverage of an elongated object, like a soldier.

Reply on news.ycombinator.com