>>> Alternatively, if you have a good set of people already in charge, 
>>> interviewing and making hiring decisions, you are not going to employ 
>>> any BS managers (or programmers) in the first place.
>>
>> I haven't seen that happen yet, but I live in hope.
> 
> Funny, the companies I've worked at, plus the customers I've dealt with 
> all seem to work this way.  I guess the ones that are rubbish don't get 
> much business, or if they do nobody comes back!

Apparently I'm applying to the wrong people...

(Actually, that's not true. I'm not applying to *anybody*. :-( But 
that's another story.)

>> Like I said, I think the difference is that in something abtract like 
>> programming, it's easier to pretend that you know what you're doing. 
>> If somebody asks you to build a wall and you can't, it's pretty 
>> obvious. ;-)
> 
> Personally I think it's pretty obvious if someone asks you to write a 
> program to show the first 1000 prime numbers, and you can't.  Even a 
> non-programmer would recognise that it wasn't working.

Sure, you'd think. But if you try to run it and it fails, you can spout 
some gibberish about "oh, it's a system incompatibility" or "yeah, it's 
a known bug in the new GCC release" or "ah yes, I'm used to working with 
RedHat. Under SuSE there's a known glitch with glibc that can sometimes 
make this happen" or any number of other BS excuses.

And, if the person doing the interviewing is some PHB, they'll sit there 
and be like "wow, this guy really knows is stuff!" (And they'll also 
completely ignore the tech dude sitting next to them going "uh, sir, 
this guy is talking utter crap!")

>> Ooo, that's harsh, man! Some of those are wicked-hard...
> 
> Yeh I'm just getting back in to doing some more of them.  Funny how just 
> 6 or 12 months later you come back and some that seemed wicked hard are 
> suddenly easy!

It's all about using the right abstractions. ;-) Some of the hardest 
programs of science and mathematics turn out to actually be trivial to 
solve once you already know how to solve it.

>> I'm trying to think of a language you could actually throw at me that 
>> I don't know, but then I realised it's actually not hard: C, Perl, 
>> Python, PHP, Bash, Ruby, VB, C#, F#, Erlang, any of those would fit 
>> the bill. o_O
> 
> I guess it also depends on what sort of job you want the person to do.  
> Do you want them to be a C++ code-monkey for their entire career at your 
> company, or are you looking for someone to do a bit more.  This will 
> determine whether you simply test their C++ skills or find some other 
> way to test them with unfamiliar languages and problems.

And that of course depends on what your company does. If you have a 
deployed application with 82 MLOC written in C++ and you just want 
somebody to maintain it, you'd better hope they know their C++. If on 
the other hand you want somebody to throw together a few housekeeping 
scripts, flexibility is probably more important.

And if you want somebody to design and build mathematical models, what 
programming languages they know is going to be utterly irrelevant - but 
you'd better hope to God they have a good grasp of algebra!

Post a reply to this message