I think I'm encountering a similar bug as described in this thread:

In my case, maptext is responding to the color var differently based on whether the object's icon has an unnamed state. When the icon_state is set to "", maptext is colored one way if a "" state exists in the .dmi, and another way if it does not.

atom/icon = 'icons.dmi'
maptext = "<FONT COLOR=WHITE>test"
icon_state = "white"
icon_state = "black"

for(var/turf/T in world)
T.color = rgb(rand(50,255),rand(50,255),rand(50,255))
for(var/turf/T in world)
T.icon = T.icon == 'hasblank.dmi' ? 'icons.dmi' : 'hasblank.dmi'

for(var/turf/T in world)
T.icon_state = ""
T.icon_state = initial(T.icon_state)

You'll need icons.dmi to have the states "white" and "black", and hasblank.dmi to have "white", "black", and no name. Or just download the zip below. Changing the icon while the states are all "" will also change the colors!! Is there something I'm totally missing for why this would happen or is this still a bug?


Bumping this because now I am finding that maptext isn't affected by atom.color at all. (The old behavior with the presence of a no-name state changing the color is no longer occurring.)

I tried the above test with the font tag, and without the tag by setting the map control's default text color to white instead. Either way, text remains white regardless of the turf color. To me this is actually more convenient but I wanted to ask if this was an intentional change or not.

(BTW, when using the classic interface, maptext defaults to black, whereas in the interface editor it defaults to white.)