Hello,
Has anyone considered something like HDNes but with the added capability to replace (groups of) tiles with 3D models? That's what I had originally planned for Nesalizer, but since I seem to have a hard time getting around to it, I thought I'd just throw the idea out there.
The implementation I imagine is having a bunch of contexts, where a context consists of at least a context identifier and an (x,y) position on the screen. A context could be defined by a particular tile, by a tile but only if surrounded by particular other tiles, or by something more complex. With that general framework, the original graphics, replacement graphics, a 3D model, etc., could later be drawn for each context. The context identification/generation code would be decoupled from the drawing code (except perhaps for passing stuff like tile data if available), so that most of the work would turn into just identifying different types of contexts after a while. It would probably be possible to do fancy stuff like detect how far into a level a game has scrolled and generate different context identifiers depending on that.
Having an editor would be nice too, where you could e.g. select a few tiles and drag-and-drop them to have the editor generate a .png with a filename that uniquely identifies those tiles and allows them to be edited.
It might seem silly, but I've felt pretty bad over the last few years about not getting around to this, and in a way it would give me some closure if someone else implemented it.
Has anyone considered something like HDNes but with the added capability to replace (groups of) tiles with 3D models? That's what I had originally planned for Nesalizer, but since I seem to have a hard time getting around to it, I thought I'd just throw the idea out there.
The implementation I imagine is having a bunch of contexts, where a context consists of at least a context identifier and an (x,y) position on the screen. A context could be defined by a particular tile, by a tile but only if surrounded by particular other tiles, or by something more complex. With that general framework, the original graphics, replacement graphics, a 3D model, etc., could later be drawn for each context. The context identification/generation code would be decoupled from the drawing code (except perhaps for passing stuff like tile data if available), so that most of the work would turn into just identifying different types of contexts after a while. It would probably be possible to do fancy stuff like detect how far into a level a game has scrolled and generate different context identifiers depending on that.
Having an editor would be nice too, where you could e.g. select a few tiles and drag-and-drop them to have the editor generate a .png with a filename that uniquely identifies those tiles and allows them to be edited.
It might seem silly, but I've felt pretty bad over the last few years about not getting around to this, and in a way it would give me some closure if someone else implemented it.