> Heh. I don't even go that far.

Yeh the written test (which they do unsupervised) is meant to test the basic 
Engineer (and maths) fundamentals.  In the mechanics test we throw in a few 
easy electronics questions and vice versa.

> I just have a bunch of questions like "why do they call it raytracing" and 
> "what's the difference between a deterministic and a non-deterministic 
> state machine" and "what causes a SIG_SEGV" and "what's the difference 
> between an inner join and an outer join" and "what's an advantage of UTF-8 
> over other encoding schemes" and "what does a translation look-aside 
> buffer translate".

We have ones like "draw a diagram to show how an injection molding machine 
works" :-)  It's surprising how few people get taught this at university (I 
certainly wasn't), yet most people at least have a rough idea what's going 
on.

> If you can't get at least a handful of such questions right from outside 
> the field you're applying for, I know you're inexperienced, unschooled, or 
> uninterested/incapable of expanding your responsibilities.

We have a separate section of the interview where we give them an LCD and 
tell them to take it apart and talk through what they think each bit does, 
why it's like that, etc.  It's a good one because virtually nobody has ever 
done this before so it puts everyone on an equal level, and there are plenty 
of mechanical and electrical things to talk about.

Post a reply to this message

> We have a separate section of the interview where we give them an LCD 
> and tell them to take it apart and talk through what they think each bit 
> does, why it's like that, etc.  It's a good one because virtually nobody 
> has ever done this before so it puts everyone on an equal level, and 
> there are plenty of mechanical and electrical things to talk about.

And here I was thinking that all the components are microscopic...

Post a reply to this message

Darren New wrote:

> Heh. I don't even go that far. I just have a bunch of questions like 
> "why do they call it raytracing" and "what's the difference between a 
> deterministic and a non-deterministic state machine" and "what causes a 
> SIG_SEGV" and "what's the difference between an inner join and an outer 
> join" and "what's an advantage of UTF-8 over other encoding schemes" and 
> "what does a translation look-aside buffer translate".
> 
> If you can't get at least a handful of such questions right from outside 
> the field you're applying for, I know you're inexperienced, unschooled, 
> or uninterested/incapable of expanding your responsibilities.

...OK, how many other people here felt an irresistablel urge to try to 
answer all of those?

Post a reply to this message

>> I will admit to having done some absurd things like that in introduction
>> to programming classes. Not for having forgotten how to solve the
>> problem, but for the challenge of finding a very bad design and the look
>> on teaching assistants faces when they have to read the code and offer
>> feedback.
> 
> That's cold.  I had a friend who used the entire McDonalds value menu for
> variable names.  Or I think I brought this up before, but I knew someone
> learning c who programmed a sudoku solver and stored the whole thing in binary,
> only the binary numbers were in decimal (e.g. 110110010.999999).  It involved a
> lot of logarithms and powers of ten.

Amatures.

Real coders define a continuation monad who's operators are all ASCII 
art, and generate spaghetti code using an epimorphism over endofunctors. 
Deobfuscate THAT! O_O

Post a reply to this message

Invisible schrieb:
> ...OK, how many other people here felt an irresistablel urge to try to 
> answer all of those?

I'd be having trouble with the translation look-aside buffer right now. 
I have a feeling that it's about cache memory, or command execution 
queue, but that's about how far I spontaneously get without resorting to 
My Friend.

But aside from that - yeah :-)

Post a reply to this message

>> ...OK, how many other people here felt an irresistablel urge to try to 
>> answer all of those?
> 
> I'd be having trouble with the translation look-aside buffer right now. 
> I have a feeling that it's about cache memory, or command execution 
> queue, but that's about how far I spontaneously get without resorting to 
> My Friend.
> 
> But aside from that - yeah :-)

TLB - convert from virtual memory addresses to physical memory 
addresses. It's basically a cache of the page mappings. (Let's face it, 
accessing RAM for every memory access to figure out where in physical 
memory the requested virtual address is would be *really* slow...)

I only know this because I read the Intel reference manuals. For no 
reason other than curiosity.

Post a reply to this message

>> We have a separate section of the interview where we give them an LCD and 
>> tell them to take it apart and talk through what they think each bit 
>> does, why it's like that, etc.  It's a good one because virtually nobody 
>> has ever done this before so it puts everyone on an equal level, and 
>> there are plenty of mechanical and electrical things to talk about.
>
> And here I was thinking that all the components are microscopic...

Err no, each component is roughly the same size as the display itself.  eg 
for a typical 2.0" phone display each component is going to be the same 
order of magnitude.  Of course the electronic circuits that are actually on 
the glass panel itself are too small to see, for the electronics guys (yes 
we only get male applicants) we just show them some simplified circuit 
diagrams of important stuff, showing them the transistors under a microscope 
wouldn't be too helpful.

Post a reply to this message

> ...OK, how many other people here felt an irresistablel urge to try to 
> answer all of those?

I could answer the first one :-D

Post a reply to this message

Invisible schrieb:
> Real coders define a continuation monad who's operators are all ASCII 
> art, and generate spaghetti code using an epimorphism over endofunctors. 
> Deobfuscate THAT! O_O

I'd first need to deobfuscate your sentence :-P

Post a reply to this message

> Real coders define a continuation monad who's operators are all ASCII art, 
> and generate spaghetti code using an epimorphism over endofunctors. 
> Deobfuscate THAT! O_O

We had a function called Viagra in our code.  It was called every second to 
check if the lifting arm on our robot was still in the fully raised 
position, and if not activate the motor until it hit the limit switch.

Post a reply to this message