Re: not enough real programmers?

[prs@gol.com (Peter Stephenson)]

On 9 Aug 1999 15:17:24 GMT, dlc@fc.hp.com (Dennis Clark) wrote:

>: > int x[10] = { ?, ?, ?, ?, ?, ?, ?, ?, ?, ? };
>: > 
>: > 1.  "Write a fast C program to move all the zeroes to the end of
>: > the array."
>: > 
>: > 2.  "Write a piece of code to clear the 3rd-lowest bit of a word"
>: > 
>: > 3.  "Write a function to print out a 32-bit number in decimal, but in
>: > reverse."
>: > 
>: > 4.  "Write a program to perform a "perfect shuffle" on an array of 52
>: > numbers." 
>
>  This goes a long way towards understanding the poor state of affairs in
>our CS world.  "Prove you're a programmer - hack out some code off the top
>of your head".  How much code out there comes from this mindset?  We should
>be asking programmers to plan a project by giving them a trivial project
>example via a user specifications description and see where that takes
>him/her.  It would be a more telling tale of what you really want.  Any
>damn programmer can learn a language and write a sort routine.  Good 
>programmers know how to design and plan.
>
>  My answer to the above four questions would be first "go find some code
>that has already been written by someone else to solve such a trivial 
>problem." 
>
>'nuff said,

A trivial project example?  Which proves they can design trivial
projects? 

My reason for asking about trivial algorithms is to see how they
approach problems and to see if they can think of different ways of
doing something given different circumstances.  

I'd like to know (really) how you also find out whether then can
design and plan.
  
The question remains: how do you tell if they are decent programmers?