[phobos] line 3250 too

[David Simcha]

I don't have access to osx.  I wrote those unittests recently, based on what
the documentation in std.math said pow should do for the special cases, and
they all pass on Windows.  Also, at the same time I did this, I only made
modifications to pow() to loosen the types it accepts, not to change the
meat of any function, so we're probably just uncovering bugs that have been
around for a long time.

Perhaps osx's C pow() function is broken like the one on linux and the
version(linux) statement in pow() should be changed to include Mac OS also.

On Mon, Aug 16, 2010 at 3:30 PM, Andrei Alexandrescu <andrei at erdani.com>wrote:

> I replaced lines 3249 and 3250 with:
>
>    version(OSX) {} else assert(isNaN(pow(one, dinf)));
>    version(OSX) {} else assert(isNaN(pow(-one, dinf)));
>
> to get std.math to unittest on osx.
>
>
> Andrei
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.puremagic.com/pipermail/phobos/attachments/20100816/7e5e6a87/attachment.html>