views:

225

answers:

7

I have been asked by one of my potential employers to undertake a brainbench test for Boost (position is for C++ programmer). I have never heard/seen a brainbench test for Boost. Brainbench's website also doesn't have any hint. I was wondering perhaps it is some sort of customized solution for employers.

My question, however is, what can I expect in this test. I am pretty good with some of Boost libraries but have no or very basic knowledge of others. Does it even make sense to test for the knowledge of a set of libraries such as boost itself?

+7  A: 

IMHO, it makes no sense at all. The Boost library collection is so large that I doubt if any one person ever uses even 20% of them. Still, on the bright side, you are highly likely to know more than the people at Brainbench do.

anon
+2  A: 

I'm highly suspicious of Brainbench. To see if it was worthwhile, I was asked to take the SQL test and I had a good passing, but not exemplary score. That's without ever having used SQL. I just used common sense and test-taking skills. That tells me that Brainbench measures test taking skills and not necessarily expertise in a particular subject.

plinth
Shhh - you're giving my secrets away...
sdg
+3  A: 

The brainbench test for regular C++ tests for memorization of components of C++. Depending on your level, it increases the difficulty as you progress. Some of the questions are down-right silly and only someone who debated the spec would probably known the answer off the top of their heads. I suspect it is similar for Boost...

As with Neil, I cannot imagine a test on Boost. Not only are there so many libraries, but they span such a large range of computing issues that I have doubts as to the utility of such a test. I would suggest you master the main components of boost that are popular: smart pointers, bind, function, lambdas, any, regex, casting, and maybe even threads. I wouldn't worry about the more specific libraries such as MPI or the math libs and what not.

Dr. Watson
+1 with Dr. Watson, this is what I am planning to do too.
A: 

Though Neil, et.al., state it very well, I feel the need to concur on my own :)

I'm not familiar with the brainbench series, but would consider any exam (outside of academia) that tests my recollection of minutiae to be essentially a waste of time. My ability to craft code isn't based on what I can exactly recollect from boundless libraries but instead what I can do with algorithms, tests, error checking, etc.

That said, a short test to verify that a candidate can do what he says is always in order. You wouldn't believe the number of 8's (on a 1-10 scale) who can't write 'hello, world' in their chosen language. Finally, if the purpose of the test is to check the breadth and depth of your knowledge within Boost, to be used as a baseline for responsibilities and future training, that COULD be a reasonable use.

KevinDTimm
A: 

In my opinion, having a knowledge of C++ is different from having a knowledge of the boost libraries. They are completely orthogonal to each other. The C++ standard is supposed to be independent of any particular library. NOT the other way around. So it sounds like your potential employer is really wanting a boost programmer who knows C++, but is possibly being deceptive (or incompetent?) and stating that they are seeking a mere C++ programmer. All of which would make me worried.

C Johnson
Most jobs say something like "C++ Programmer Wanted - must know SQL Server and be familiar with Equity Trading" - I don't see anything "deceptive" about that.
anon
+1  A: 

I've been forced to take some of these tests for previous jobs. They are multiple-choice questions.

The results seem to depend on more than just the correctness of the answer you give.

I think that the test takes into account how long you take to answer each question. Possibly also the number of times you change your answer.

DanS
A: 

In my experience, that test is useless to test your skills and knowledge.

However, they are a quick and cheap way to find out if you just added a buzzword to your resume or if you have actually used boost without waisting an hour of a good engineer to interview you.

For you that means if they do not actually interview you on the subject later, be careful about joining them. Companies with bad hiring practices usually makes other mistakes in treating their employees as well.