Recently i had a choice between two modules, Date::Manip, and Date::Pcalc (a perl version of Date::Calc). I originally went with Manip, but someone pointed out that manip was supposed to be slower, so i went out and benchmarked them, and came up with:

use Benchmark;
timethese(10_000,{
    'Pcalc'    => sub {
    require Pcalc;
    my $date=20020428182540;
    my @date;
    push @date,substr($date,0,4,'');
    for my $i(0..int (length $date)/2){push @date,substr($date,0,2,'')
+;}
    pop @date;
    my @diff=reverse Date::Pcalc::Delta_DHMS(@date,Date::Pcalc::Today_
+and_Now());
    return "[$diff[0]s:$diff[1]m:$diff[2]h:$diff[3]d]";},
    
    'manip'    => sub{    
    require Manip;
    $main::TZ='EST';
    my $d1=Date::Manip::ParseDate(20020428182540);
    my $d2=Date::Manip::ParseDate('today');
    my $err;
    my $delta=Date::Manip::DateCalc($d2,$d1,\$err,1);
    my $first=substr($delta,0,1,'');
    my @d=reverse split /:/,$delta;
    return ($first eq '-')?
    "[$d[0]s:$d[1]m:$d[2]h] $d[3]d-$d[5]m-$d[6]y":
    "This was posted in the future. Bummer. Odd aint it?";}
    })
[download]

Which gave me the following results:

Benchmark: timing 10000 iterations of Pcalc, manip...
     Pcalc:  2 wallclock secs ( 1.98 usr +  0.00 sys =  1.98 CPU) 
@ 5050.51/s (n=10000)
     manip: 49 wallclock secs (49.10 usr +  0.00 sys = 49.10 CPU) 
@ 203.67/s (n=10000)
[download]

Now, even realizing that manip is supposed to be slower, Thats a rather extreme difference. Anyone else have useful module benchmarks to share in the context of choosing one over the other?