in reply to Perl Exam?

Oh boy. You hit a nerve here, probably unintentionally.

If you were hiring a gardener, would you would find someone acceptable who doesn't know the difference between annuals and perenials? If you were hiring a bank security guard, would you waive the drug test so that you could save a dollar an hour off their pay? If you were hiring a grade school teacher, would you neglect a background check for sexual misconduct? If you were hiring an electrician, would you pick up some guy on the street corner who graduated high school changed flourescent bulbs before? No, you wouldn't, not unless you're very short-sighted, negligent, irresponsible, or stupid.

Some attitudes towards computer programmers:
These attitudes are based on myths. Trying to follow these myths will frustrate them and will harm progress towards your goal.

I've seen it again and again - someone who should know better hires the wrong guy (or even the right guy) because they find some kid who fits their stereotype of the computer genious. Good vibes all around, everyone is excited about the project. After a month or two of no apparent progress or a few prototypes that don't look like what you want, things start to get tense. You start to doubt your guy because you have lingering doubts about your ability to pick someone. You know you could be being taken advantage of, so you strugle every day with the question of whether this kid is taking advantage of you. You turn up the heat on him, demanding more and faster, to keep him from getting away scot free with bilking you. Eventually, he quits for a better job, or you cancel the project or more often, fire him and bring in someone younger, leaner, and cheaper, and repeat the whole progress even more spectacularly. Meanwhile, consultants don't feel like they can tell you how error prone and grueling and exacting the software development process is because they don't want to be labeled a hack and they don't want to ruin the good vibe that each project starts out with.

If you're at the doctors with horrible pains, you don't dismiss him for an intern if he says that he will have to do surgery and it will cost a lot and there will be a long recovery. People have serious control issues when it comes to programmers - they don't understand the process, but they don't understand exactly how little they understand the process (see the bullet list of dillusions above), so they tug on the reigns when they should be staying the heck out of it.

Yes, I'm getting around to an answer to your question. I just needed to post some background to the problem first, since the scope of the problem seems to have been missed.

The solution is (if you haven't guessed) is to:

On the first point, someone who is active in the community, has gone through college and studied advanced topics while they were there, has their name spread all over the Internet when you search google for it, and/or was already writing software before going to school is what you're after. This is a better test than some quiz in some ways, but people lie and play the system, so a good minimum compentency test is helpful too. Brainbench is too easy to cheat on. Reading part of perlfaq (perldoc.org) and quizzing them from that verbally should do the trick.

After you find this person and put them to work, leave them alone. Get a second opinion from a trusted source if you must - and more people should do this far more often - but don't harrass them out of fear or boredom.

Perhaps I'm overly cynical, but I've seen far more mismanagement, aborted efforts, amaturely written code leaving a time bomb that always goes off sooner or later, and time and energy wasted on politics, power plays, and second guessing.

I hope this helps. I'm sure there are other good suggestions in other notes in here - this doesn't attempt to second guess them.

-scott

In Section Meditations