Description
It would be nice if we were given more control over pixel offsets by expanding on what options are available via appearance_flags. To that end, I am proposing 8 new flags that would be available for use:

RESET_PIXEL_X
RESET_PIXEL_Y
RESET_PIXEL_W
RESET_PIXEL_Z

INVERSE_PIXEL_X
INVERSE_PIXEL_Y
INVERSE_PIXEL_W
INVERSE_PIXEL_Z

RESET_PIXEL_X/Y/W/Z basically allows for any appearances to ignore the pixel offset values of the parent atom.

INVERSE_PIXEL_X/Y/W/Z works in the opposite direction. Instead, it ensures that the parent ignores any pixel offsets, and that the pixel offsets only applies to overlays, images, or vis_contents that are added to an atom's appearance.

Why Is This Useful?
I know what you all are thinking. With a bit of mathematics, these features can already be accomplished. Nonetheless, adding these features could streamline a lot of computation.

Reset Pixel
Example 1: Let's say you want a simple shadow system that shows where your character is location on the x/y axis, even if they're flying. As they go up in altitude, you would normally need to calculate how far they are in the air, and adjust the pixel offsets for the shadow. With RESET_PIXEL_X/Y/W/Z, that computation could be ignored.

Example 2: Visual effects, like particles could benefit from having this feature available. For example, let's say you want to make an effect that animates the character around a particle, or above -- or however, then this becomes useful.

Inverse Pixel
Example 1: The main scenario where having this flag available would be useful would be for turfs. It's possible that you may want to have certain turfs able to have visual accessories, some that may need to be offset to work properly. As it stands now, the only viable solution would be to attach a dummy object to the turf(s)with the appropriate offsets attached to that dummy object, then applying any visual accessories to that object.

That's wasteful. With INVERSE PIXEL we can bypass that waste.

Example 2: Another use here would be in managing UI. For example, with INVERSE_PIXEL, it allows you to simulate modifying all over/underlay by changing a single value which can be useful. Although, I suppose in this case, a dummy object would be sufficient (albeit less efficient if this was available, I believe.)


Anyway, just a thought. Let me know if the idea sucks or not.