pouët.net

Hyde, don't assume that the image will be stretched to fullscreen. It was true even a few years ago, but GPUs are increasingly using aspect preserving stretch and letterboxing if the screen mode doesn't fit the monitor aspect. And it's what eg. the BP projection system does.

Assuming square pixels and having sane defaults (like: choose the desktop res as default or at least another mode that has the same aspect ratio) that the user can just click through is way safer nowadays.

Scamp already posted the link where I tried to explain all that shit above anyway ;)

thanks, kb !

Awg, this is awful stuff. But, assuming square pixels is easier than getting those fractions right, so I guess, in the end, it´s good news.

I guess there aren´t any way of figuring out if the driver stretches or not... (?)

I've dug into the respective Windows APIs for two days and haven't found any definite, authoritative way of figuring out what happens when you set a certain mode. No chance.

you could, in theory, find the corresponding registry entries for the drivers, but that can differ wildly

that's what i meant by "no definite, authoritative way" ;)

eww. I´m not programmer enough to go to those lengths :)

Hyde: Back to matlab with you, theory-boy

sloomy: pff. I may have hygienic issues with the registry, but you know I wanna be your dog. JAG VIL VARA DIN HUND!

Hyde: all that mess was easier back in the 68000 days, when we were younger (especially you) :)

There was shit to be handled back then and there is new shit to be handled now. Machines - I just don't like them.

Hyde: In fact, "scheisse" is what you really don't like! :D

Navis: 2.35:1 for TVs is in the pipeline