It always amuses me when people choose to put down a language purely upon speed or based on an arbitrary benchmark. A recent node (Use Perl wisely, not cleverly) based its put down of Perl on a variety of reasons. Now ignoring the fact that the arguments used against Perl can just as easily be applied to Java and ASP (speed and memory) I do wonder why programmers or wannabe programmers get so hooked on speed and ram. Perhaps its a fast car thing that you grow out of ;).
There have been a handful of projects over the last five years that have required any kind of speed in their run. Aside from the fact that I used Perl in each case, I simply made use of the hardware and software (webserver, database etc) correctly and tried to use the best algorithm for the job.
Now I am fortunate that any intensive program I have written has had database, network and internet latency to compete with so the speed comparisons really become moot.
The point? Well a project I had to complete a while ago had to compete with a C++ multi-tier network system. The solution, replace with Perl. Why I hear you cry. Why use something that is so (tonque in cheek mode activated) bloated, slow, old, doesn't do (insert OO principle here) and hasn't had 6 O'Reilly books published on it in the last month alone?
Well, because we used it cleverly. Development time was reduced, debugging time went down from 3 weeks (or more!) to no more than a day, Perl was easy to teach to a larger audience and we were within 10% of the original C++ code speed.
So whats your point
Well my point is this. I'm glad your language has (insert feature here) and if I have to, I'll learn it and use it and do so happily.
But time and time again Perl has gotten the job done quickly enough and has allowed me to go home on time and have a life.
Think about that the next time you are still debugging a memory leak at 2am.
An afterthought
I do believe Perl can do most tasks, however I do know when and when not to use it (cleverly remember?). I just don't believe the speed arguments are justifiable.
There have been a handful of projects over the last five years that have required any kind of speed in their run. Aside from the fact that I used Perl in each case, I simply made use of the hardware and software (webserver, database etc) correctly and tried to use the best algorithm for the job.
Now I am fortunate that any intensive program I have written has had database, network and internet latency to compete with so the speed comparisons really become moot.
The point? Well a project I had to complete a while ago had to compete with a C++ multi-tier network system. The solution, replace with Perl. Why I hear you cry. Why use something that is so (tonque in cheek mode activated) bloated, slow, old, doesn't do (insert OO principle here) and hasn't had 6 O'Reilly books published on it in the last month alone?
Well, because we used it cleverly. Development time was reduced, debugging time went down from 3 weeks (or more!) to no more than a day, Perl was easy to teach to a larger audience and we were within 10% of the original C++ code speed.
So whats your point
Well my point is this. I'm glad your language has (insert feature here) and if I have to, I'll learn it and use it and do so happily.
But time and time again Perl has gotten the job done quickly enough and has allowed me to go home on time and have a life.
Think about that the next time you are still debugging a memory leak at 2am.
An afterthought
I do believe Perl can do most tasks, however I do know when and when not to use it (cleverly remember?). I just don't believe the speed arguments are justifiable.
Back to
Meditations