Admittedly, there’ll be problems no matter what if they’re not testing their program with even slightly-modified configurations. But I’d rather encounter a piss-poor, ugly interface and blame it on the programmer not being able to draw (see Everett Kaser software) than one that completely breaks simply because I like the color blue more than grey or I like to sit more than 10 inches away from the monitor. If your software makes me open up the control panel, I’m clicking ‘add/remove programs’ before ‘display properties’.

I hear this all the time. It’s certainly the common wisdom. Is it really true?

One aspect I tried to suggest in the discussion of UI but I didn’t say explicitly is that the inputs required for the game impact how much UI design one should be doing. Obviously, if all the inputs for a game are expected to come from a gamepad — left/right/up/down/shoot — then there’s no real need to make the game look (OR feel) like Windows. I think, however, that games that require you to use a mouse for pointing and clicking activities present a stronger case for using a more consistent look and feel. I know that everyone _thinks_ gamers will reject games that use standard GUI elements, but I’ve not seen any actual evidence of that, since so few have tried. Have you ever met anyone who thought that Spaceward Ho! (to pick one example) was less fun because it used standard MacOS GUI elements?

Lastly, there’s a middle ground between “use ugly Windows dialogue boxes and textareas for everything” and “rewrite the mouse tracking code, only get it slightly wrong, driving the user insane.” As you point out, correctly, probably the most important thing is that the apps _act_ like the users expect, regardless of how they look.

I’m going to wave my hands at the DirectX performance issues and say that if solving them is important, they can and will be solved. I’ll also mutter something about Moore’s Law under my breath.