Strategic View

[Modo44]

That is somewhat gratuitous logic.

In what way?

Decreasing grid size to approach or equal pixel (or memory location) size is effectively removing the underlying grid from the human's perspective.

I agree. I't doesn't make the actual grid go away, though. And it isn't really doable at the moment due to lacking computing power. Graphics mapping is way simpler than computations on such detailed data. Ask anyone working with digital maps.

[Sir Og]

Decreasing grid size to approach or equal pixel (or memory location) size is effectively removing the underlying grid from the human's perspective.

That doesn't make a lot of sense. That's like saying there are no microbes because you can't really see them and you should not worry about them.

Even if you make the grid very small it is still a grid with all the issues associated with it.

[wodan11]

Okay, let me answer with a question: what are the problems with a small (pixel size) grid (yes, rectangular)? I think, if we enumerate them, we'll find they aren't problems at all.

[Modo44]

Okay, let me answer with a question: what are the problems with a small (pixel size) grid (yes, rectangular)? I think, if we enumerate them, we'll find they aren't problems at all.

On current hardware, you'd just swap distortion issues for major performance issues. But yeah, once that is solved in X years, the problems associated with large squares/hexes could be mostly gone (specifically: hidden deep enough not to matter).

[wodan11]

So, you assume that it's not feasible to use a X,Y coordinate system without downgrading system performance, and thus your conclusion is that it's not a viable solution?

As long as the new coordinate system did not drastically add more elements (which it would not, as the number of buildings/units in play would remain the same), I think the conclusion does not logically follow.

[Modo44]

If you make a very detailed map, the quantitative data (i.e. what's contained in what area, distances, routing) will also be very detailed, creating a serious computing requirements hike. Remember that you still have to count what the land provides before you apply building bonuses, and the map won't create itself. At this detail level it won't matter much if you base the underlying map on hexes or not, since that problem will be easy to deal with.

If you choose to ignore the additional data to not increase requirements, then making the map more detailed won't make sense in the first place, and you'll be back to what Civ 5 brings: Big map "dots" (i.e. hexes in the most sensible implementation).