gdiplus: Implement GdipCreateRegionRgnData. Take 2.

[Vincent Povirk]

Dmitry,

I agree about the underscores. It seems like a very minor point, and I
personally wouldn't have expected it to be a problem.

>From my perspective, it seems like adding FIXME's should be easy,
they'd be potentially helpful for diagnosing problems, and there
doesn't seem to be a drawback to adding them (either they're never hit
and have no effect, or they are hit and they indicate a potential
problem). I really don't understand your objection to adding them, and
that is what makes your response seem "defensive" to me at least. But
yes, if your patch is accepted without FIXME's I will send a patch to
add them myself when I have time.

I really don't feel like this is anything new. If you're a Wine
developer, you will sometimes have to make changes you disagree with
or think are unnecessary to get a patch in. Sometimes this is to
address very minor points. You will also have to decide which feedback
is really important and which you can ignore or argue with, and then
you have to update your model when it turns out Alexandre doesn't see
things the way you thought he would. I don't know if this is a good
thing. Maybe it means we're losing some good contributions due to
people getting frustrated by the process and giving up. But I've found
this to be the case for at least as long as I've been a Wine
contributor, and since you've been around longer than I have, I'd
expect you to be used to it by now.