Yukihiro Matsumoto wrote:
> Hi,
> 
> In message "Re: [ ruby-Patches-6323 ] BigMath::log can be made much faster"
>     on Thu, 26 Oct 2006 08:32:39 +0900, <noreply / rubyforge.org> writes:
> 
> |BigMath::log is slow for numbers greater than about 10, and can be made much faster by exploiting log identities. I submitted an RCR for this a year ago, and C Erler graciously optimized my reimplementation, but it looks like it died on the vine. Here's C Erler's message to ruby-core:
> 
> The maintainer, Shigeo Kobayashi, refused to apply this patch, since
> he thinks this kind of optimization should be done in application
> level.  I'm no good at mathematics, so that I had no way other than
> trusting him.  
> 
> 							matz.
> 
> 
Somewhere buried in my collection of books are some test suites for math
functions. I can probably find them either in my library or on the web
somewhere if you want to have another run at this. There's absolutely no
point in doing it *without* tests, though.