Re: [5/5] WineD3D: Fix fogstart == fogend with table fog

[H. Verbeet]

On 21/08/07, Stefan Dösinger  wrote:
> Feel free to suggest a better way to increase the fogend. My way works, but it
> is not the nicest way.

Does this really matter? To me, this looks like a rather ugly hack to
get the DCT to shut up.

[5/5] WineD3D: Fix fogstart == fogend with table fog

[Stefan Dösinger]

Am Dienstag, 21. August 2007 13:08 schrieb H. Verbeet:
> On 21/08/07, Stefan Dösinger  wrote:
> > Feel free to suggest a better way to increase the fogend. My way works,
> > but it is not the nicest way.
>
> Does this really matter? To me, this looks like a rather ugly hack to
> get the DCT to shut up.

Well, it is d3d's (ugly) definition of a corner case, in the same way gl has a 
definition of that corner case(d3d has two, actually). Of course we can wait 
for a game to require this behavior, but will we recognise it as such? Or 
will we spend a lot of time for debugging, only to find out that it is a 
well-known problem we deliberately didn't fix?

Agreed, some problems are hard/impossible to fix, so we shouldn't fix them 
just to make the DCT happy, like flat shading or D3DRS_WRAPx. But this fog 
fix is pretty simple.