[PATCH v2 4/5] gdiplus: Fix hatch brush patterns that require anti-aliasing.

[Jeff Smith]

On Sun, Jun 14, 2020 at 9:15 AM Esme Povirk (they/them)
<vincent at codeweavers.com> wrote:
>
> On Sun, Jun 14, 2020 at 2:16 AM Jeff Smith <whydoubt at gmail.com> wrote:
> > > Why not blend them as ARGB?
> >
> > What are you suggesting that is not being done now?
>
> Use blend_colors. It's intended to blend two ARGB colors. If it
> doesn't work correctly, fix it.

Thank you for clarifying.  I just want to make sure that I am
addressing the right concerns.  As you suggest, I will simply use
blend_colors for now, and leave further blending improvements out of
this patchset.  I guess I'll also make the test accept inexact matches
for now.

> If the color values are slightly off from what native produces, but
> there's no clear reason ours are worse, I don't think we should care.

If we have an alternative that matches what native produces, and
there's no clear reason it is worse, perhaps we should.

> Matching the output perfectly is not necessary for application
> compatibility.

True, but I think this is an area where small errors can compound quickly,
so I do think fidelity with native is warranted here if practical.

> I don't see a reason why blending should work differently in this case
> from other cases that use blend_colors.

I agree that it shouldn't in principle, and I hope that holds true in practice.