On Fri, Apr 5, 2013 at 9:41 AM, Rico Schüller <span dir="ltr"><<a href="mailto:kgbricola@web.de" target="_blank">kgbricola@web.de</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
On 05.04.2013 09:13, Nozomi Kodama wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
s = 0.75f;<br>
- if ( order > 2 )<br>
- s += 5.0f / 16.0f;<br>
- if ( order > 4 )<br>
- s -= 3.0f / 32.0f;<br>
+ if (order > 2)<br>
+ s = 1.0625f;<br>
+ if (order > 4)<br>
+ s = 0.96875f;<br>
s /= D3DX_PI;<br>
</blockquote>
<br> Doesn't the compiler do that already?</blockquote><div><br></div><div>Most likely,since those are literal constants.</div><div>Besides, 5.0f / 16.0f is more readable than a pre-computed result.</div><div><br></div>
<div>So I don't think this patch is really useful.</div><div><br></div><div>Frédéric</div>