sam fisher
Well-Known Member
I just realised my version is clocked at 240
Well, that's Fenix's little gift, isn't it?sam fisher said:I just realised my version is clocked at 240
Good points but not really feasible, sry.rokdcasbah said:i wonder if there's a way around sdl's slowness when alpha-blending, say by doing a lookup table of sorts.
like if there's 10 different color circles, most of the time they'll be the right color, but when they overlap you could detect where the surfaces collide and change the color to what's in the lookup table. obviously there'd be a lot of things to lookup, but the fewer colors there were, the better, and you could ignore combinations of, say, 4 or more. also you'd have to use a solid color background (but maybe you could again do a lookup table or fixed point addition for that).
i have no idea if that would work or work fast enough but i'm just throwing it out there...maybe i'll code it up myself (in 70 years)
oh, just wait 70 years like i said, then we'll see who's "sry", java-man.Nagelfar said:Good points but not really feasible, sry.
rokdcasbah said:oh, just wait 70 years like i said, then we'll see who's "sry", java-man.Nagelfar said:Good points but not really feasible, sry.
Yeah, probably on the GP3000 Looking forward to it.
i use alook up table to cheat at anglesrokdcasbah said:i wonder if there's a way around sdl's slowness when alpha-blending, say by doing a lookup table of sorts.
like if there's 10 different color circles, most of the time they'll be the right color, but when they overlap you could detect where the surfaces collide and change the color to what's in the lookup table. obviously there'd be a lot of things to lookup, but the fewer colors there were, the better, and you could ignore combinations of, say, 4 or more. also you'd have to use a solid color background (but maybe you could again do a lookup table or fixed point addition for that).
i have no idea if that would work or work fast enough but i'm just throwing it out there...maybe i'll code it up myself (in 70 years)
sam fisher said:i use alook up table to cheat at anglesrokdcasbah said:i wonder if there's a way around sdl's slowness when alpha-blending, say by doing a lookup table of sorts.
like if there's 10 different color circles, most of the time they'll be the right color, but when they overlap you could detect where the surfaces collide and change the color to what's in the lookup table. obviously there'd be a lot of things to lookup, but the fewer colors there were, the better, and you could ignore combinations of, say, 4 or more. also you'd have to use a solid color background (but maybe you could again do a lookup table or fixed point addition for that).
i have no idea if that would work or work fast enough but i'm just throwing it out there...maybe i'll code it up myself (in 70 years)
How do you connect angles and the alpha blending?! *confused*