Hi!
Greycat: There have been quite a few posts pointing out how horribly unrealistic and game breaking the large sprawl areas are. Then there are the complaints about the broken migrants. The following is the first clear example I found going backwards, but there are more discussing the percentage of land taken up and the reasoning of cities claiming the entire river they are built at. In other threads, I have even seen people calling 31.13 buggier than the first 31.01 release:
I wouldn't really call DF updates patches (for one thing, they don't overwrite the old version), but in any case nothing good happened to fortress mode (aside from a few bug fixes), it's harder to find a spot to embark on, worldgen takes longer, and it's even harder to find quests in adventure mode now that the human towns got replaced with an experimental update. Personally I'd stick to 3.12 until all the kinks get worked out.
There has been quite a bit of discussion here.
Rexfelum: First of all, I recommend that you mess around with world gen using the editor within the game until you got used to (except when copying world gen information from others). This way, you get more information about what the settings mean and also what their limits are.
After the initial settings about world size, you get the variance and range settings for the individual factors. Each is a set of four numbers: The lowest number the characteristic can have on your world map, the highest number it can have, how much it may vary between two tiles along the x-axis and how much it may vary between two tiles along the y-axis. In general, unless you want to create an extreme world for experiments (like a world so hot that all living things evaporate on it), you should not touch the minimum and maximum values in those initial sets. They are the normal ranges you want to have.
For the variance between neighboring tiles, values somewhere between 200 and 500 are usually save and non-complicated. Once you get below 100 with the variance, it means that you have very little chance for difference between neighboring tiles, resulting in lots of rejects as there is not enough variance on the map.
The rule of thumb for the variance settings is: The smaller the values (=little difference between neighboring tiles), the bigger the subregions (forests, plains,...). Think about it. If two neighboring tiles differ only slightly in their height, for instance, a mountain chain is quite likely to exist. But if two neighboring tiles could differ drastically in height, there can be steep cliff faces, resulting in things like the lone mountain.
There is one very important and useful exception to the rule about subregions, though: Volcanism. While volcanism influences what stone layers can be there and of course allows the presence of volcanoes, volcanism does NOT disrupt subregions. Therefore, unless you have any special preferences about volcanism, I recommend
maxing out the variance for volcanism as you can then have volcanism influences in many different types of terrain, allowing for a greater variety of terrains to experiment with.
After those sets, you get the mesh sets, a feature I have not experimented with. So, no information from me on that.
Next comes the setting for the features. Their influence on the rejects and the maps differs:
Minimum mountain peak number is a rejection condition: mountain peaks require great height AND great difference between the peak and the surrounding tiles. If you set the elevation variance too low in the initial sets I have explained, you will not get a single mountain peak and if you have minimum peaks set to more than 0, you will only get rejects as there are not enough mountain peaks.
The next two settings for oceans basically influence the start of the generation, I think. If you pick 1 one or more partial edge oceans, I think any map generated will have one ocean tile placed at its edge in the beginning. I don't think I ever saw a rejection because of that. Likewise, complete edge oceans also never created rejects for me, so I assume that it causes one (or more if you select more than one) entire edge to be automatically set to ocean tiles at the beginning of world gen.
Minimum volcano number is actually a misnomer: The game will place exactly as many volcanoes on the map as you state there - provided areas of high volcanism are available.
I skip explaining the titan and demon settings as they are more a flavor thing, although a higher number of titans will probably increase the risk of civs getting extinct.
Next come the good and evil squares. Once the general map is generated, some tiles are turned good and some are turned evil and these settings let you specify how many. There are three settings for each to allow you to specify what kind of areas you want. For instance if you have many desired evil squares in small subregions, then you should get a lot of small groves or single mountains that are haunted/terrifying or whatnot. If you have many desired evil squares in large subregions, then you don't get those small groves spread across the map but rather a large forest spanning many tiles filled to the brim with evil. Note, however, that a large subregion has more tiles than a small one, which is why the values increase over the settings.
As good and evil squares restrict settlements of races, you can think about whether to increase their number in order to increase the locations where races can't settle or you can decrease them so as to have more space for races to expand.
The next settings are the terrain count settings which I was talking about referring to the forest, oceans, unusables and so on. Basically, these are rejection settings. That is, the game generates a map and then checks whether there are as many tiles of each region as you specify there. These values do not cause such regions to be generated (that is the job of the variance settings), so if you set the maximum height to be nearly identical to the minimum height and the require more than 0 mountain region tiles, you will never get a world as no mountain tiles are generated and the check fails.
These settings are among the most useful for you as they allow you to clearly tell the program that you want to have at least so and so many tiles of desert or forest or whatever. However, as each of these parameters is checked, you should nullify all things you do not particularly care about (after all, let's say you want a world with lots of forests and mountains; then you don't want to have a perfect candidate rejected because it has 999 desert tiles instead of 1,000 ).
Erosion cycle and the two river settings are fun settings for landscaping. Genning rivers, the game does the following: First, it selects as many springs as you specify with the pre-erosion setting (if there are not enough potential locations, you get a reject). Then, these springs are connected to the edge of the map or an ocean by a river following the rules of height, of course. Then, for each erosion cycle count, the rivers dig themselves deeper into the terrain - think of the grand canyon as an example of a very high erosion cycle and the local creek that is basically level ground as an example for a low erosion cycle. Finally, if the post-erosion river starting location number is lower than the pre-erosion one, the excess springs and the rivers they have created are removed again, leaving behind the dry canyons of their riverbeds.
Periodically eroding extreme cliffs is a setting which I recommend to set to "No". It is meant to make sure that you only have slopes and never steep cliffs, but usually, the game does not create much terrain that is problematic anyway, so having the occasional cliff is a good option.
Orographic Precipitation is a factor of reality. Unless you are designing and extreme world, I recommend you set it to yes, but in the end, it is just a matter of taste.
Maximum Number of Subregions is an important rejection setting. Unless you want to filter out worlds with small subregions, I recommend you ALWAYS max out that value. Basically, it says how many different mountain ranges and forests and plains there can be in the world. If your terrain is very diverse, you may have too many physical subregions and then this will cause a reject.
This is followed by the cavern layer settings. If you want to experiment with all the cavern layers, you should keep the cavern layer number to three. The min/max settings afterwards are important for setting what the caves look like and the magma and bottom layer are the "bonus layers".
The z-level settings that follow are again important for performance and difficulty: The more z-levels in total, the slower the game will probably be in the end game when you have cleared all those tiles. However, the more z-levels you have above layer 1, the more digging and building you can do before encountering the first cavern layer and thus the first mandatory battle.
The minimum/maximum natural cave size, mountain cave and non-mountain cave and cave visibility settings are about the additional caverns where kobolds, legendary beasts and the like settle. For the minimum/maximum, I recommend minimum 100 and maximum 500 so that the kobolds have enough space to retreat to. The number of caves influences how difficult it is for civs to survive the first years - if you increase the number of caves, the chances of civs meeting monsters early on increases resulting in civs that get butchered before they get enough population to survive.
Number of Civilizations is problematic in the current version. You should set it to 16 to 20, I think, for medium regions and adjust it accordingly for other map sizes.
Playable civilization required should always be set to yes as it means that at least one dwarven civ is generated (though it need not survive).
Below that, there is a number of minimum values: Set all of them to 0! They are rejection conditions that are not needed for 99% of your world designs. They only cause otherwise perfect worlds to be rejected because of some minor issue, so get rid of them. All the things you want to influence are usually already set by the other settings.
I hope this kind of helps you a little bit in your attempts to mess around with world gen.
Deathworks
Author
Topic: Dwarf Fortress 0.31.13 Released (Read 115987 times)