[Freeciv-Dev] Re: (PR#7287) Extended Topologies

["rwetmore@xxxxxxxxxxxx" <rwetmore@xxxxxxxxxxxx>]

<URL: http://rt.freeciv.org/Ticket/Display.html?id=7287 >


Don't make the mistake of assuming tiling is equivalent to making arbitrarily
complex border wraps or necessary to deal with a sperical reduction to 2-D.
Tiling has the additional property of needing simple(r) wraps so that one can
scroll over the surface consistently with a rectangular GUI window.

This is why unrolling the complex wraps until one has a super-tile that now
tiles simply in all directions is a useful trick to identify valid cases and
see what the effects actually are.

It may actually be a good implementation optimization, as one can capture all
the complex math of extended wraps in a limited region and simply tile these
regions with no more complex math than the current basic 8 topologies, i.e.
splitting the problem into two steps may make for a lot of common code with
hard math relegated to cachable lookups.

Thus tiling is far from absolutely necesary if you are willing to live with
dual hemispherical views or other ways of presentation.

It is also why I think that only four sided tiles are useful in reducing a
spherical surface to a 2-D flat plane. If you give up on tiling, i.e. don't
reduce things to simple wrapping, then other tile shapes can work, but they
will not be trivially scrollable in the GUI canvas.


Cheers,
RossW
=====


Jason Short wrote:
> <URL: http://rt.freeciv.org/Ticket/Display.html?id=7287 >
> 
> rwetmore@xxxxxxxxxxxx wrote:
> 
>>I think that tiling is the additional operation that Jason is adding into
>>his requirements, which is nice but not necessary, and not something
>>considered in the original model above.
> 
> No, tiling is absolutely necessary to get the spherical effect. 
> Otherwise you can cross the equator at South America but you can't do so 
> at Africa!
> 
>>I suspect that only the square tile shape permits tiling in 2-D as Jason
>>would like, and since this is the quincunx model, it is the only one
>>that can produce a nice 2-D tiled map. Can anyone confirm this?
> 
> As I said before the triangle would also permit it.  But this is useless 
> to us.
> 
> jason