What's the algorithm for deciding in which base to express the post count?

[jasong]

Is it random? And what are all the possible bases for this particular algorithm? I mean the bases that are being used right now, not ones you could potentially add.

All the bases seem to be a power of 2, base 2, 4, 8, 16, 32, have I missed or forgotten some that don't fit this pattern?


edit: Oh yeah, also there's post counts where they factor the numbers. Maybe the ones in a weird base are all prime?

[Stargate38]

I think it depends on the subforum, or possibly even the thread number (i.e. this thread starts with 446485). It could also be a random seed-based algorithm.

[Xyzzy]

This code is really inelegant, but it works.

$zzz = $post[posts]; 
$xxx = $zzz; 
if ($xxx == 0) { 
    $zzz = rand(1, 10000); 
    $xxx = $zzz; 
} 
$tmp = ""; 
$choice = time(); 
if ($choice % 20 == 0) { 
    $listofradices = array("2", "8", "10", "16"); 
    shuffle($listofradices); 
    $radix = $listofradices[0]; 
    $tmp = strtoupper(base_convert($xxx, 10, $radix)) . "<SUB>" . $radix . "</SUB>"; 
} else { 
    if ($xxx == 1) { 
        $tmp = "1&times;"; 
    } 
    for ($iii = 2; $iii < $zzz; $iii++) { 
        $counter = 0; 
        while ($xxx % $iii == 0) { 
            $counter = $counter + 1; 
            $xxx = $xxx / $iii; 
        } 
        if ($counter > 1) { 
            $tmp = $tmp . number_format($iii) . "<SUP>" . $counter . "</SUP>" . "&times;"; 
        } else { 
            if ($counter == 1) { 
                $tmp = $tmp . number_format($iii) . "&times;"; 
            } 
        } 
    } 
    if ($xxx == $iii) { 
        $tmp = $tmp . number_format($xxx) . "&times;"; 
    } 
    $tmp = substr($tmp, 0, -7); 
}

[Dubslow]

Looks like 1 in 20 chance to get binary, octal, decimal, or hex post count, otherwise (19 out of 20) it's shown in factored form.

[retina]

What about ...

for ($iii = 2; $iii * $iii <= $zzz; $iii++)

Fewer wasted clock cycles = more P95 clock cycles.

[Antonio]

Originally Posted by retina

What about ...

for ($iii = 2; $iii * $iii <= $zzz; $iii++)

Fewer wasted clock cycles = more P95 clock cycles.

Sorry - ignore me, just being stupid today

Or ...

    for ($iii = 2; $iii < $xxx; $iii++) {

and do away with the multiply.

[retina]

Originally Posted by Antonio

and do away with the multiply.

If $xxx is a prime then we get back the original algorithm.

ETA: But it could be this:

for ($iii = 2; $iii * $iii <= $xxx; $iii++)

[Antonio]

Originally Posted by retina

If $xxx is a prime then we get back the original algorithm.

ETA: But it could be this:

for ($iii = 2; $iii * $iii <= $xxx; $iii++)

But you also need to change ...

    if ($xxx == $iii) { 
        $tmp = $tmp . number_format($xxx) . "×"; 
    } 

to ...

    if ($xxx != 1) { 
        $tmp = $tmp . number_format($xxx) . "×"; 
    }

or it misses the final prime.

[axn]

Nothing new under the sun

[Mark Rose]

I'm still waiting for

if (!($xxx & $xxx + 1))
    $tmp = 'M' . log($xxx + 1, 2);
elseif ($choice % 20 == 0) { 
...