Basically there are three ways of doing a minimap:
1) each tile is represented by a single colored pixel
- pro: "real" minimap (like you expect it to be)
- contra: if there's multiple objects on a tile only one can be displayed, example: creature from Player A fighting creature from Player B on territory from Player C
2) a "satellite" camera (a far away "tv screen" of the original game)
- pro: probably the most easy one to realize code-wise
- contra: what is it good for? just zoom out the regular camera and you have the "minimap" on the whole screen
3) a complex map (each tile is represented by an area of pixels, movable icon overlays are displayed on top)
- pro: most informative, most fancy one
- contra: extremly hard to code, needs alot of new artwork, probably strongly CPU consuming (as long as we don't manage to solve the current performance issues, we shouldn't introduce new ones)
I vote for 1) and optionally 3) for whoever feels doing it in the far future (I won't do this hard task, at least not now
).