Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Captain Mayday]

I've been pondering about how to add more variation to creatures, given the current RAWs, and I came to ponder that perhaps using tags with variables more than once (say DAMBLOCK:5 and DAMBLOCK:10 on the same creature), that perhaps it would randomise between them.
I was also thinking the same technique could be used with ETHICS to have more varied cultures.

Can anyone verify if this will actually work?

[KaelGotDwarves]

I'm not sure myself how variation would work under the current system outside of attacks, prefstring, sex, and personalities. That's why we're waiting for the update with the next DF version.

One workaround I tried for a mod of mine was having different human civ entity_default entries.

All used HUMAN creatures but each civ had different ethics, word preferences, weapons/armor/items- stoic, order/valor worshipping knights, oceanic-partying pirates with looser morals, evil/death demon worshiping humans with no INTELLIGENT and CAN_CIV + CAN_LEARN instead (everyone will declare war on them for failure to communicate).

So civ entries already allow for a wide range of cultures. Toady already announced he was going to switch how ethics work in the future... See thread below for Toady's input on how ethics work.

http://www.bay12games.com/forum/index.php?topic=22151.0

[Captain Mayday]

Yeah, I've got the same workaround. I'm really interested in allowing for a single entity entry to have varied ethics and such amongst it's spawned civs.

[Footkerchief]

I want to see varying civ ethics too.  Teldin in that linked thread posted this:

That's a decent idea, how about a random percent for each one? Something like:

[ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:N:0:100]

The 0:100 would be a simple percent chance, you could add multiple varieties, like:

[ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:N:0:90]
[ETHIC:MAKE_TROPHY_SAME_RACE:ACCEPTABLE:N:91:100]

Thus you have randomness between civs; a 9% chance that a particular civ, when genned, will accept it. It would be a bit difficult this way though as I think individual civs don't have seperate ethics variables, just the overall civ type (ie. all dwarves share the same ethics, currently).

which seems like a decent start (ignoring the fact that his probability math is a little screwy).  However, it doesn't allow for the fact that you'd want various "evil" traits to correlate with each other.

[i2amroy]

To answer your question, currently if you have two of the same tag on a creature, (with the exception of prefstring) the second tag will overwrite the first one and the first one will have no effect.

But it would be cool to have random values in certain things.

[Granite26]

I've always been a fan of arithmatic modifiers on a geometric scale for computer based ratings.

In English:  Each factor is -5 through +5, but the effects kick in at different scales, so +5 for the first effect, +10 for the second, +20 for the third, +40 for the fourth. (this is just an example, implementation details are not my concern)

That way, Civs, Adventuring Groups, Families, Governments, Individuals and Races all get individual says on what a person thinks of Human Skull Totems.  Obviously the government (civ) currently controls what the laws are.  (I REALLY want to see goblinoids failing to live in a human settlement because they fail to understand why murder is wrong).

The geometric scale means that you don't have to worry so much about making sure that all the numbers for each element add up to something, or that each contribution is only 1/X of the max.  Doing the numbers right keeps evil civs evil, but make everybody's favorite dark elf a natural possibility.

Also, Lawfulness should be a stat/trait.  I've met people who consider something immoral purely because it's against the law.  I've also met people who value breaking laws just because they're there, and the rare individual who does what they feel is right regardless of the law.  Most (humans), I feel, slightly are slightly lawful.  They'll obey laws they agree with, expect other people to obey laws they agree with, but will easily break laws they strongly disagree with.

[Squeegy]

So humans are lawful neutral?

[Jay]

[ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:N:0:90]
[ETHIC:MAKE_TROPHY_SAME_RACE:ACCEPTABLE:N:91:100]

[ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:90]
[ETHIC:MAKE_TROPHY_SAME_RACE:ACCEPTABLE:10]
Makes a ton more sense.

[Footkerchief]

[ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:N:0:90]
[ETHIC:MAKE_TROPHY_SAME_RACE:ACCEPTABLE:N:91:100]

[ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:90]
[ETHIC:MAKE_TROPHY_SAME_RACE:ACCEPTABLE:10]
Makes a ton more sense.

Agreed.  Optionally this could be compressed into a single line: [ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:90:ACCEPTABLE:10]

And possibly further adjusted to use ratios rather than percentages, meaning they don't have to add up to 100: [ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:6:ACCEPTABLE:1]

[D_Malachi]

Could also go one simpler and drop the names of ethic ranks/reactions and stick with straight numbers, such as...

[ETHIC:MAKE_TROPHY_SAME_RACE:X:X:X:X:X:X:X:X:X:X:X:X:X:X:X:X]

...wherein each field represents one reaction in ascending level of dislike, from Acceptable to Unthinkable, and ideally implemented as weighted values/ratios. This would contain a field for every known reaction level.

PROs:
Simpler to have a generic overview of all types of reaction.
Probably easier to handle code-wise.

CONs:
That's a LOT of values to handle, the engine might not like it.
Possibly a simple matter to mess up on the modding end of things.

[Footkerchief]

^^^ Yeah, with mine including the names, the tag could get VERY long.  On the other hand, without the names it's cryptic and easy to mess up, as you said.  I can't think of a really elegant solution.

[Jay]

^^^ Yeah, with mine including the names, the tag could get VERY long.  On the other hand, without the names it's cryptic and easy to mess up, as you said.  I can't think of a really elegant solution.

[ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:90]
[ETHIC:MAKE_TROPHY_SAME_RACE:ACCEPTABLE:10]

or

[ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:6]
[ETHIC:MAKE_TROPHY_SAME_RACE:ACCEPTABLE:1]

Hmm?  I don't see any cons with that..  Sure, it might make several lines per ethic, but entity entries are huge now anyway.

[i2amroy]

^^^ Yeah, with mine including the names, the tag could get VERY long.  On the other hand, without the names it's cryptic and easy to mess up, as you said.  I can't think of a really elegant solution.

[ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:90]
[ETHIC:MAKE_TROPHY_SAME_RACE:ACCEPTABLE:10]

or

[ETHIC:MAKE_TROPHY_SAME_RACE:UNTHINKABLE:6]
[ETHIC:MAKE_TROPHY_SAME_RACE:ACCEPTABLE:1]

Hmm?  I don't see any cons with that..  Sure, it might make several lines per ethic, but entity entries are huge now anyway.

I like this one the best.

[D_Malachi]

Yeah, that might be the optimal method for ethics to work - it's simple and allows for the most versatility. Easy to add or remove possible ethical reactions easily with minimal modding experience. Very simple to read, made even easier if it works on weighted values or ratios, as opposed to a cumulative percentage. Very tough to screw up, and it would probably be a rather simple matter for the game engine to handle, I'd assume.

The method I chose would be a pain in the neck, though it's possibly the simplest as far as vertical or horizontal scrolling are concerned. A non-modder, or anyone without an at-hand "cheat sheet" of the order of reaction levels, could easily create a race that doesn't work properly, if at all. Compact, but at a cost - confusion and many problems. Probably not many engine-related problems, as far as I can see.

The method Jaybud proposes, I don't really see that as working that well from a technical standpoint, unless the engine can handle a dynamic amount of values in a single bracket. I know it can handle "optional" values, such as equipment tokens in the entity files. These can have a "quality" value, and is not required to complete an entity. However, having a semi-random amount of fields might be a problem. I don't doubt its merit, though, it's certainly a simple matter of adding new values, simply add two values to an ethic field, but it might be outside of the engine's capabilities. If it would work, only Toady would really know the answer to that, it would definitely be a viable alternative to the current "static" method of doing it.

[LegoLord]

I tried making a sword with burn and slash damage.  It didn't work, so instead I just gave it straight up burn damage. Even when it was set to [DAMAGE:60] with a 133% material modifier, it was still a brutally efficient combat weapon.

Boy, I got really distracted with this post.