Kaioken wrote:

No, it wouldn't, you didn't understand. You'd have the DMI data, but you wouldn't be able to intuitively actually access the icons themselves, i.e. isolate each graphical parts. That would need to be implemented in the language, like you can already isolate an icon_state and dir for instance with new /icon(file,state,dir), you'd need some method (probably the same) to isolate animation frames and whatnot.

I posted the new proc they would need to add, and what it would return; with that information you could determine anything you wanted: Total duration of an animated state, the frame delay on any direction of any frame, or whatever else... What are you having trouble understanding? You can already graphically isolate and cut out specific frames of specific icon_states if that's what you're claiming is impossible, even though you seemed to post that yourself.

Jeff8500 wrote:

I made a program that saves/loads similar maps, and it only took about two hours to write. My format is really similar, except it only supports one turf, area, obj, and mob per tile ('cause I was lazy, and that's all I needed).

A "similar format"? Unless the system can read a .dmm, and save as a .dmm, its not doing what it needs to be doing. I've written map saving/loading systems before, doesn't mean they're compatible with DM; which would be the goal here.