Re^4: Why should I use perl 5.10?

Faster? Anyone got some numbers to prove this?

use Benchmark;
timethese(100000, {
'Test' => 'eval{die "x" x 100000 if $] < 5.01}',
});
[download]

Under perl 5.8:

Benchmark: timing 100000 iterations of Test...
      Test: 15 wallclock secs (14.98 usr +  0.00 sys = 14.98 CPU) @ 66
+73.34/s (n=100000))
[download]

Under perl 5.10:

Benchmark: timing 100000 iterations of Test...
      Test:  0 wallclock secs ( 0.02 usr +  0.00 sys =  0.02 CPU) @ 62
+50000.00/ (n=100000)
            (warning: too few iterations for a reliable count)
[download]

Whooooo ... how good is *that* !!!!!

Cheers,
Rob

Comment on Re^4: Why should I use perl 5.10? Select or Download Code

Replies are listed 'Best First'.
Re^5: Why should I use perl 5.10? by Anonymous Monk on Aug 07, 2008 at 19:29 UTC
In a less trivial example... I was benchmarking a module I wrote against one I found on CPAN solving the same problem. I decided to see how using 5.10 compared to using 5.8 at the same time. My module (which was already faster) saw a 10% improvement with 5.10. The CPAN module saw a negligible improvement (about 1%) using 5.10. using 5.10 didn't hurt and in some circumstances added a nice boost. Depends on how you are using perl. Maybe later I'll dig a little deeper and see what the CPAN module did differently to get so little benefit out of 5.10.	[reply]

Replies are listed 'Best First'.

Re^5: Why should I use perl 5.10?
by Anonymous Monk on Aug 07, 2008 at 19:29 UTC

I was benchmarking a module I wrote against one I found on CPAN solving the same problem. I decided to see how using 5.10 compared to using 5.8 at the same time.

My module (which was already faster) saw a 10% improvement with 5.10. The CPAN module saw a negligible improvement (about 1%) using 5.10.

using 5.10 didn't hurt and in some circumstances added a nice boost. Depends on how you are using perl.

Maybe later I'll dig a little deeper and see what the CPAN module did differently to get so little benefit out of 5.10.

[reply]