FE Map Creator (also generates random maps)

Resources Toolbox
1

/copy pastas from SF

Fig A: a generated, random map

An FE tailored map creator, that can also generate random tiles to fill in empty space in your map or create an entirely random map. The generation rules can be edited, too. It also allows locking specific tiles, such as a castle or mountain range, that the generator should not modify and merely build around, as well as drawing in generic terrain types, which it will turn into fitting tiles of that type during generation:
-why doesn’t this board have spoiler tags-

Fig B: pin some tiles in place, let the program generate the rest around them


Fig C: miracles
-/this is unacceptable-

Supports opening and saving .mar files for inserting the maps into roms, and can open Tiled files. It can also create an approximation a map based on an image and the tileset used, if you just have a picture and not the data file.
The vanilla GBA FE maps and tilesets, as well as map generation rules for each tileset, are included in the download so you can immediately jump into map making.

Special thanks to Pi for ripping all the GBA FE maps, which I referenced to generate the initial tileset rules.

FE Map Creator download zip

If the above two links don’t work for you, go to the main site → FE Map Creator

You’ll need .NET Framework 4 to run it, if the program doesn’t work you probably need to download that:
.NET Framework

13 Likes
2

I am excite

3

hmm, now this seems much more workable, for random fun shit

4

Source code pls.

5

is it just me, or does this thing literally create random maps? because i’m trying to get an organized map and this is just trolling me

6

that’s literally its purpose.

on a side note i feel this can be fun for certain tilesets and really helpful aswell, but other tilesets provide little to no help, and are simply fucking fun to look at.

on a side note, yeti if ya hear this, the pintile function for the dessert maps don’t work, and the program won’t save maps in tiled format propperly (as in the tilesets don’t load) though those both might simply be internal problems i am having, but i thought i would mention it anyways.

7

Aww, don’t be a Luddite.

But anyway we could at the least train it better. Probably using a larget training set for it. So mappers keep mapping because you’re teaching the mapgen

8

why would a mapper be working on making a map generator when they already know how to map and have no intention of making it automated

and still working on “stuff”

1 Like
9

That’s not ‘being a Luddite’, and I object to the mischaracterization.

That’s power of design. Procedural design is great in a pinch for say, a D&D game when a GM may be pressed for time, but if you’re going so far as making a map for a hack, I assume you have some sort of vision of how you want a chapter to play and flow. No one at IS pressed a generate map button and out was Cog of Destiny spat.

Also, you cannot code aesthetic sense and an artistic eye into a map generator - If any of you somehow are at this level, then I ask what you’re doing not applying your talent in a more marketable fashion.

2 Likes
10

Er, sorry, I meant that half-jokingly.

But in terms of actual map design, it can be used as a randomizer, but its chief use is to allow you to set constraints, i.e. “This tile must be a forest, this a plains, these mountains” etc. So you can plan out that and allow the tool to come up with the aesthetics of the map. iirc from that Yeti has told me it just works by predicting what tile fits a square best based on how IS has used them in their maps, i.e. it has learned from a data set.

1 Like
11

I’ve actually demonstrated that it can make very decent indoor maps if you plan out how to set terrain tags, use sticky tiles for specific things like thrones and stairs, and edit the results manually. Sure it takes work, but even if it never competently does indoor well, it makes very convincing outdoor maps. Like, the whole point made in Prime’s mapping tutorial is “nature is random” so what could be more random for the outdoors than a random map generator?

12

except it’s not random map generator random; nature follows a system of sorts
rivers don’t become mountains
forests are rarely on the beach
etc

13

i’d argue that the map generator would be more useful for a level designer because you can set the flow manually and let the “flavor” (paths, etc) and the “noise” be generated for you, then just fix the specific terrain for yourself

2 Likes
14

If you have to go through all that work to get a somewhat decent generated map, might as well just do it by hand anyways since you’re already halfway there.

15

Tbh Yeti’s map generator is pretty awesome, and even if you’re an experienced mapper, you could still use its maps as a base or as some source of inspiration.

16

idk the experienced mappers keep saying we have pretty much no use for it
but sure we can just ignore that

17

The experienced mappers also haven’t even tried using it for a decent period of time?

“Hey we have this vaccine that can cure influenza”
“NO FUCK OFF MY MOTHER’S CHICKEN SOUP MAY TAKE 4 WEEKS INSTEAD OF FOUR DAYS BUT I’D RATHER USE IT THAN YOUR SO-CALLED ‘SCIENCE’!”

18

Not really comparable.

People keep saying “Oh, you can use it for inspiration or just fix the errors”.

  1. I get my inspiration from elsewhere, not FE maps (IE I look outside or something just pops into my head).
  2. Why do I want to fix the mistakes some program pops out? I’d rather fix my own mistakes.
19

You can specify how maps generate too! Look at sticky tiles which prevent specific tiles from changing, and using terrain tags to get a more variable effect when you generate.

20

Klok that’s not even a comparison.

You want the random map generator to be a vaccine, sure. But since you know, it’s random and most of the time unhelpful, let’s make it for an unknown disease.
What mappers would be would be people who are using a vaccine they created and know works 100% on the disease they’re trying to treat(the map) vs taking a gamble with a random vaccine that might cure the disease faster(I’ve tested it out, it never helps me), but 99% of the time it will muck up everything and provide no results or results that the mapper would have to recreate a different string of the original vaccine to cure.

If someone obviously has no knowledge to create their own vaccine, then something that could work would seem great. But if someone has a vaccine they already knows works perfectly fine, why would they bother with spinning that wheel of chance.