I recently came across a task that originally pitted Java programmers against Lisp programmers, in support of the premise that, for a given task, Lisp programmers produce more concise (less bugs) solutions more quickly than their Java couterparts. I'm not concerned with discussing the premise, nor the validity of the test methods which were all done to death on slash-dot a couple of years ago.

What does interest me is seeing what Perl programmers make of the task. It isn't possible to assess the time taken to produce results in this environment, so the assesment process will be:

• Accuracy of results.

  If you do not match the sample output, you lose.

• No of lines of executable code.

  Counted manually. Putting three statement for line won't reduce your score. This isn't golf!

• Time taken to process the sample dataset (on my machine).

  The results must be produced by processing the file. (Ie. Pre-processing the input and just producing the output will be cause for disqualification!).

• A 'maintainablility quotient'.

  A tough one this. I'm not really sure of a good way to measure this?

  My thought is that I will produce an assessment, with justifications for each response awarding (negative) scores for each point if unmaintainability. Then, if there is enough interest, we can discuss it. (Someone else will have to assess my entry: Volunteers?

The scores will be the product of the above three values (lines * seconds * maintainability-demerits), with the lowest score being the winner.

Assessment will begin on Wednesday. (Arbitrary cut-off date that could be extended if need be.)

The challenge

All the information required to complete the challenge is detailed here in the first four bullet points.

Have fun!