Re: d3dx9_36 [patch 6/6]: Implement D3DXSHEvalHemisphereLight

[Rico Schüller]

Well I think, if you rework a series, please add something like try xyz 
or so ... Also try to avoid to add / remove some patches, at least if 
they actually don't have anything todo with the rest of the patch set 
(if they could be submitted on their own).

On 27.09.2012 10:00, Nozomi Kodama wrote:
> +    TRACE("%u, %p, %p, %p, %p, %p, %p\n", order, dir, &top, &bottom, rout, gout, bout);
> Sure, that's an other form of the trace. I think we hadn't one with such 
> a version... other traces in the file look like (the one marked with * 
> is my preferred version):
> TRACE("out %p, order %u, a %p, b %p\n", out, order, a, b); *
> TRACE("(%p, %p, %u)\n", pout, pin, n);

I don't think it makes much sense to print the address for &top. It 
doesn't gather some useful information, does it? Instead you may print 
the actual content for r,g,b if you like.
> +    if ( gout )
> Also here, I prefer "if (gout)" ... I know there are several
> occurrences for the other style ... I should have mentioned that earlier 
> ... it's not something which is critical...

git grep "if ("  dlls/d3dx9_36/* | wc -l reveals ~2000
while git grep "if ( "  dlls/d3dx9_36/* | wc -l does ~100
> +                    ok(relative_error(expected, blue_out[j]) < 
> admitted_error,
> +                      "Blue: case %u, order %u: expected[%u] = %f, 
> received %f \n", l, order, j, expected, blue_out[j]);
> Please indent 8 instead of 2 spaces when a new line is used ...

Cheers
Rico