A new map script (MountainCoast)

[[to_xp]Gekko]

a couple suggestions:

1) you may want to borrow from Erebus the code that deletes forests/jungles in the first ring of tile around the starting settler ( unless it's elves : I'm pretty sure Erebus already does this ) . It's really good for AI, allows them to start off better and we all know how critical that is. it also makes jungle starts a little less crippling, which is definitely nice ( possibly remove jungles in the starting BFC instead of just the first ring? )

2) you may want to check out cephalo's PerfectWorld mapscript ( http://forums.civfanatics.com/showthread.php?t=239982 ) : the heightmaps there are awesome, it really gives the land a realistic feeling ( it communicates altitude very well ) . I'm saying this cuz in Mountaincoast sometimes the mountain ranges feel kinda "out-of-place" . I love how they look, but sometimes they seem misplaced. ( for example when you see a huge mountain range right next to the coastline... it kinda gives a "umh, it's pretty weird that there's a mountain range HERE" feel

hopefully this is helpful to you. this mapscript is great already, and it can only get better. keep up the great work!

 

[Deon]

for example when you see a huge mountain range right next to the coastline...

But the script is called MOUNTAINCOAST )

Also:

(Well, it looks like this place in Civ should have mountain->hill->coast transiton though... So I get your idea).

 

[lonkero173]

The forests in desert: I think I forgot to change a constant to a variable when I made the script more customizable. This most likely only manifested in the newest version (since I changed the desert moisture limits). If the above is correct this will be fixed in the next release.
The flood plains: No, its not intentional, the script just places flood plains in all desert tiles with access to fresh water. I don't know how to test for a river....

@JDexter:
I know, and its been causing loss of hair at an early age... However since this is quite rare and mostly a graphical issue I decided to give my hair some time to grow back before I give it another shot.

@Gekko
2) I can put in some code to guarantee that peaks are not directly next to water but I consider this rather pointless (why not have the occasional peak next to water?)(ever heard ofnorway? )
1)This is definitely going in now that you mention it, I'll get it in when I manage to get the starting position system working.

And also thanks for hanging around and giving feedback, this script would still be pretty
unplayable without it (this goes for the rest of you people as well )

 

[Jean Elcard]

The flood plains: No, its not intentional, the script just places flood plains in all desert tiles with access to fresh water. I don't know how to test for a river....

Don't test for fresh water. There is a method called isRiver() available for CyPlot objects. Use this one instead. It'll do the trick.

 

[Bartholomew]

I'm not giving more feedback because I'm too busy enjoying an Epic Game on the previous version of your script. It's one of the most enjoyable games I can remember, in large part because of the "feel" of the map.

And I agree, you need mountains right next to the ocean! I really like the narrow coastal shelf thing your script generates sometimes.

 

[Verdian]

And I agree, you need mountains right next to the ocean! I really like the narrow coastal shelf thing your script generates sometimes.

I agree, I just wish I didn't keep starting there... It took me 4 re-loads to not be in a sliver of land surrounded by ocean and/or mountains.

Still, once I got set up somewhere nice, the map was beautiful and a lot of fun to explore. I just felt bad for the AI that could not reload. Although the AI Lanun did amazingly well like that and stretched their Empire along half of the coast.

 

[[to_xp]Gekko]

any plan to add marshes and scrubs to the mapscript in the next version? they add some more much appreciated variety to the land

 

[JeffSteel]

Marshes are actually rather common on the coast near where a river enters into the ocean. Until there's a reliable way to improve them somehow, I would say the marsh level is fine as is. A few more scrubs couldn't hurt though.

 

[Rod]

its beautiful !!

 

[Psychic_Llamas]

lonkero173, this map scipt has gotten absolutely awsome since i last tried it! great to see youve fixed so many of those bugs right now my only concern is the civs flavoured starts. i was lanun and started in a massive plain with a 2 square lake, and my neighbours were illians... in the desert...

other than that (which i bet you are getting around to fixing) its excelent! great work!!

 

[lonkero173]

I just got the first version of the starting position system working to a satisfactory (not very good yet but improving) degree. I also fixed some minor issues (marsh on hills, river check for flood plains...), as well as changing the marsh distribution a bit. Scrubs will also be present.
I will redo part of the algorithm to calculate local yields instead of just look at terrain types, and I'll get it to deal with player seperation a bit better.
The starting positions are not altered in any new way.

@JanusTalaiini:
Confirmed the issue with mana nodes and my resource system(my memory sucks), next version should habe this fixed (not 0.9.5, need some sleep now)

File is attached to the first post as usual

 

[[to_xp]Gekko]

sounds great, thanx man! now go get some well-deserved sleep

I'll give it a try and report back in a while.

 

[[to_xp]Gekko]

feedback on 0.9.5. :

1) some civs start very near to each other, but you said that's gonna be looked at soon so it's ok.

2) some "rough" shapes can still sometimes be generated. I had a land bridge between 2 lakes that was exactly 2 tiles wide and 6 tiles long, fully covered with forests. it really looked like a bridge, kinda weird

3) are totally isolated valleys supposed to be possible? I had a small valley with some fur and mana in it, but no one started there. don't know if this is intended. it was on the coast though, so indeed reachable by water, just not by land.

4) sometimes valleys near to mountain ranges are entirely made of hills. maybe they could be "flattened" a bit so that they're still very hilly, just not 100% ? aside from being pretty much impossible to settle a decent city there, it looks a little weird.

5) I noticed forests are still present in the starting spots, so I guess you haven't added the "forest - clearing" code from Erebus yet?

6) is there a way to place some code to prevent goblin lairs ( and maybe other lairs ) from spawning right next to a starting plot? it can be kinda hairy with those nasty guys as neighbours, you know

7) cosmetic: "normal" forests often spawn on tundra/snow instead of "snow-covered" forests. nothing serious, but it would be nice to have trees be appropriate to the climate.

keep up the great work!

 

[Psychic_Llamas]

talk about fast! thanks lonkero (Downloads now)

 

[lonkero173]

1) This is the biggest problem with the system, and part of the reason I'm starting a rewrite about now. However do note that these nearby spawns are on decent starting terrain, and should they be moved further apart, they will end up in some marginal terrain.(I tried to get a balance between "too close" and "miserable place")

2) Randomness.... not much I can do about this unfortunately. (wish I could)
3) No, thats what the water is for. Considering water accessible makes things much easier now, and especially when I eventually get to the warping version.
4) Do you have any good ideas how to detect such areas? I could do some random terrain lowering but anything sufficent to make a difference to such areas would also wreck havock in the "succesfull" areas. Also areas with plains/hills arent really meant for settlement, that one of the reasons for the size of the maps.
5)

The starting positions are not altered in any new way.

(thought I'd just push this out as it is and gather some comments, I'll get to this soon)
6) I don't know. Probably yes but I don't know how yet. (will try to fix, low priority)
7) Will be fixed

 

[Bartholomew]

On (5) above, "clearing" the space around the start position: my vote is NO. I like the feel of having to literally hack my civilization out of the wilderness, and I like having my environment shape my tech choices (say, bee-lining bronzeworking) rather than automatically going for what my civ is good at.

 

[Valkrionn]

On (5) above, "clearing" the space around the start position: my vote is NO. I like the feel of having to literally hack my civilization out of the wilderness, and I like having my environment shape my tech choices (say, bee-lining bronzeworking) rather than automatically going for what my civ is good at.

I would tend to agree, especially with forest clearing on mining now... not quite so bad for the AI anymore.

 

[[to_xp]Gekko]

so, can we settle on a dropdown option? it'd be good to experiment with that setting ON and OFF, to see what benefit the AI can reap out of that.

regarding 4) , I am no coder but maybe you could check for 4*4 ( and bigger ) areas that are 100% hills, and then randomly flatten some tiles?

 

[lonkero173]

Dropdown it shall be then, its not much of a job adding more.

4) yes, simple but should do the job. However, the thing with checks like this is that they tend to add up. Doing even this properly (seperate check at each square) will require (4*4)*(mapX*mapY) operations (4*4*60*40=38400 operations for standard map size)(and python is on the slow side of languages) just to find the areas, and possibly more than that actually flattening the terrain. For most cases this makes little difference as most people dont care if a map script takes a few seconds longer to run. However, I do, since I tend to generate the maps in the dozens. (current total is probably in the hundreds) Thus I am somewhat reluctant to add lots of little things like this that add fairly little to the quality of the map while producing extra work both as a coder as well is increasing generation times.
(I'll do something similar nevertheless, next version...)

Also I'm making decent progress on the starting position system, it now gives much better evaluations of the quality of a potential starting position. Still going to improve the way it handles player separation, maybe this evening.
As a side note: the script will come in two files from now on, this significantly eases my work.

 

[[to_xp]Gekko]

you could ask Jean Elcard and cephalo if they can think of any quicker method, if they can help I'm pretty sure they will

question: why 2 different files?