php, python, perl, again.

Brian Moon brianm at
Fri Jun 30 19:48:25 UTC 2006

> You - send me 3 *command*line* scripts reproducing the
> spike (below) + 3 logs that those 3 command line
> scrips produce on your machine.

Perhaps I should change the name of my report then.  Since my need is 
not on the command line, I am not testing the command line.  I am 
testing how these languages, within Apache, perform with memcached.

> In worse case that work could imply installing debian
> on some fresh box and/or doing some research in the
> way perl/python/PHP integrate with apache. I might do
> that work for you too, but not for free. However, if
> the spike will be preserverd in command line scripts,
> then it should be easy to detect.

Again, how those languages integrate with Apache is part of this test. 
So, I don't really see the validity of removing it from the equation.

> If your benchmark does *not* claim that PHP does 5-10
> times better than Perl / Python on small 'gets' from
> memcached, then this is going nowhere, please
> disregard this email.

I have no claim.  I have my test results.  Its not my opinion or my 
belief.  I have asked for anyone that knows perl or python better to 
help me improve their test cases so that the test results are as even as 

However, you have sparked my curiosity.  So, I ran the tests.

Again, unless I have mistakenly made perl or python inefficient, PHP 
still looks just as fast.

I want to be clear.  I am not rooting for one over the other.  In fact, 
we had hoped that python would be the winner in our office.  We like the 
language and kind of want to start using it more.  But, we won't be 
using it for this project.


Brian Moon
Its good to be cheap =)

More information about the memcached mailing list