![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
What the hell?
I've got hyperbolic interpolation now... (well, mostly...)
I had given up on getting that to work hours ago, and was resigned to just using linear (incorrect) interpolation.
However, when writing the manual, I noticed something that seemed fishy in my linear interpolation part of the manual, compared to what I had in the code. Since it didn't seem right I checked the code, where I learned that I am ever the absolute fucking dumbass as usual. Due to a copy/paste error from my x interpolant function to my z interpolant function (copy/pasting for the new interpolants), some values were reversed and it was totally screwing stuff up. A FUCKING COPY/PASTE ERROR.
So, now, just like that, I have hyperbolic interpolation. It still doesn't fully work, as part of the texture kind of jumps at a specific point. And there are still some unrelated issues such as the rasterizer mucking up some of the triangles, and the clipper still not correctly clipping the shadows to the far and near planes (though it's no longer doing that weird thing where it would wrap to the right/left if the shadow went off the left/right edge). But hot damn, I'm much better off now than I was before I started writing the manual.
Anyway, the gist is, due to a stroke of blind, stupid luck, I'll probably get at least 15-20 (I hope) out of the 30 points for implementing hyperbolic interpolation now instead of the 0 out of 30 I would have gotten for just settling for linear interpolation. Or, if you want to rephrase it, due to a stroke of blind, stupid (bad) luck, I wasted many hours trying to figure out why my hyperbolic interpolation originally wasn't working, when it was just a simple matter of variables being switched around incorrectly. DUE TO A COPY/PASTE MOTHERFUCKING ERROR.
On the bright side, I managed to not lock myself out of my condo this time, at least, so there is that.
I suppose I should go back to finishing the manual now rather than posting to LJ, shouldn't I? There's still a bit to do on that and it's almost 5:00 am.
I've got hyperbolic interpolation now... (well, mostly...)
I had given up on getting that to work hours ago, and was resigned to just using linear (incorrect) interpolation.
However, when writing the manual, I noticed something that seemed fishy in my linear interpolation part of the manual, compared to what I had in the code. Since it didn't seem right I checked the code, where I learned that I am ever the absolute fucking dumbass as usual. Due to a copy/paste error from my x interpolant function to my z interpolant function (copy/pasting for the new interpolants), some values were reversed and it was totally screwing stuff up. A FUCKING COPY/PASTE ERROR.
So, now, just like that, I have hyperbolic interpolation. It still doesn't fully work, as part of the texture kind of jumps at a specific point. And there are still some unrelated issues such as the rasterizer mucking up some of the triangles, and the clipper still not correctly clipping the shadows to the far and near planes (though it's no longer doing that weird thing where it would wrap to the right/left if the shadow went off the left/right edge). But hot damn, I'm much better off now than I was before I started writing the manual.
Anyway, the gist is, due to a stroke of blind, stupid luck, I'll probably get at least 15-20 (I hope) out of the 30 points for implementing hyperbolic interpolation now instead of the 0 out of 30 I would have gotten for just settling for linear interpolation. Or, if you want to rephrase it, due to a stroke of blind, stupid (bad) luck, I wasted many hours trying to figure out why my hyperbolic interpolation originally wasn't working, when it was just a simple matter of variables being switched around incorrectly. DUE TO A COPY/PASTE MOTHERFUCKING ERROR.
On the bright side, I managed to not lock myself out of my condo this time, at least, so there is that.
I suppose I should go back to finishing the manual now rather than posting to LJ, shouldn't I? There's still a bit to do on that and it's almost 5:00 am.
