Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  
Pages: 1 [2] 3

Author Topic: Making more diverse genetics  (Read 12420 times)

noob

  • Bay Watcher
    • View Profile
Re: Making more diverse genetics
« Reply #15 on: February 19, 2011, 02:36:54 am »

can i suggest sterilization? i need a master race of cows with with incredible quantities muscle and fat. skinny cows get in my way...
Logged
LOSING IS FUN!

Demicus

  • Bay Watcher
  • The formless enigma
    • View Profile
Re: Making more diverse genetics
« Reply #16 on: February 19, 2011, 03:30:25 am »

can i suggest sterilization? i need a master race of cows with with incredible quantities muscle and fat. skinny cows get in my way...
Kill off anything that doesn't meet your standards before they have a chance to taint your breeding stock
Logged
All shall embrace the unquenchable flame
Dwarf Fortress: The weak shall be culled, so the strong can have nicer socks.

noob

  • Bay Watcher
    • View Profile
Re: Making more diverse genetics
« Reply #17 on: February 19, 2011, 07:11:23 pm »

can i suggest sterilization? i need a master race of cows with with incredible quantities muscle and fat. skinny cows get in my way...
Kill off anything that doesn't meet your standards before they have a chance to taint your breeding stock
they produce milk
Logged
LOSING IS FUN!

Demicus

  • Bay Watcher
  • The formless enigma
    • View Profile
Re: Making more diverse genetics
« Reply #18 on: February 19, 2011, 10:30:33 pm »

can i suggest sterilization? i need a master race of cows with with incredible quantities muscle and fat. skinny cows get in my way...
Kill off anything that doesn't meet your standards before they have a chance to taint your breeding stock
they produce milk
Well even if sterilization is added to the game, it'll most like just be castration, not nuetering. Castration is rather simple for the time period DF is set in. Nuetering females is highly invasive and would have a very low chance of succeeding. So it might just be simpler to just butcher the indiserables
Logged
All shall embrace the unquenchable flame
Dwarf Fortress: The weak shall be culled, so the strong can have nicer socks.

NW_Kohaku

  • Bay Watcher
  • [ETHIC:SCIENCE_FOR_FUN: REQUIRED]
    • View Profile
Re: Making more diverse genetics
« Reply #19 on: February 19, 2011, 11:08:55 pm »

You only need one male to impregnate as many as 48 females.  As long as you don't already have 50 cows, cows can still get pregnant, and the 50 population cap only stops pregnancy, not birth, so you can wind up wtih 60 or even 100+ animals, depending on litter size.  Just slaughter every male but the one bull that has the most desirable traits, and slaughter every female beyond the top 48 that have undesirable traits.
Logged
Personally, I like [DF] because after climbing the damned learning cliff, I'm too elitist to consider not liking it.
"And no Frankenstein-esque body part stitching?"
"Not yet"

Improved Farming
Class Warfare

QuakeIV

  • Bay Watcher
  • Cant resist... must edit post.
    • View Profile
Re: Making more diverse genetics
« Reply #20 on: February 21, 2011, 12:01:05 am »

Iv seen people who haven't had a redheaded person in their family for quite a while, and then out of nowhere two out of three of their children have red hair.

What color was the mailman's hair?

Nice, this thread has just become educational and scarring.

Looks like a full house for this one.




Also:
Sterilization sounds totally awesome, especially the part where you have to slowly weed out the recessive genes as you identify them.
« Last Edit: February 21, 2011, 12:02:45 am by QuakeIV »
Logged
GENERATION 9: The first time you see this, copy it into your sig on any forum and add 1 to the generation. Social experiment.
I wish my grass was emo, then it would cut itself.
Quote from: Jesus
Quote from: The Big Fat Carp
Jesus, you broke the site!
Sorry, Bro.
link to quote

Thundercraft

  • Bay Watcher
    • View Profile
Re: Making more diverse genetics
« Reply #21 on: February 21, 2011, 02:12:57 am »

In The DF2010 Little Questions Thread there was a brief discussion on the new grass requirements of domestic animals:
Apparently cows go through a lot [of grass]. Currently I have 4 goats and a few truckloads of rabbits buzzing around a roughly 20x20 pasture and they are ok, but I don't know how that translates to cows.

Question: Has genetics in 0.31.19 been coded so that an usually small animal breed requires less grass than an usually large animal breed? The BODY_APPEARANCE_MODIFIER seems to affect the quantity of products from butchering, so I wondered if it affects grazing, too.

For that matter, does genetics have any measurable influence on the rate of food consumption for civilizations?

I'm really curious to find out. And I propose that this could make the basis for an excellent !!!science!!! experiment. However, I'm too off-put by .19 to try this out myself at the moment. Maybe in .20 or later...

If it turns out that this aspect of genetics has not been implemented, I think it deserves being brought up in a .19 feedback thread (considering it might affect grazing).
Logged

NW_Kohaku

  • Bay Watcher
  • [ETHIC:SCIENCE_FOR_FUN: REQUIRED]
    • View Profile
Re: Making more diverse genetics
« Reply #22 on: February 21, 2011, 10:10:28 am »

In The DF2010 Little Questions Thread there was a brief discussion on the new grass requirements of domestic animals:
Apparently cows go through a lot [of grass]. Currently I have 4 goats and a few truckloads of rabbits buzzing around a roughly 20x20 pasture and they are ok, but I don't know how that translates to cows.

Question: Has genetics in 0.31.19 been coded so that an usually small animal breed requires less grass than an usually large animal breed? The BODY_APPEARANCE_MODIFIER seems to affect the quantity of products from butchering, so I wondered if it affects grazing, too.

For that matter, does genetics have any measurable influence on the rate of food consumption for civilizations?

I'm really curious to find out. And I propose that this could make the basis for an excellent !!!science!!! experiment. However, I'm too off-put by .19 to try this out myself at the moment. Maybe in .20 or later...

If it turns out that this aspect of genetics has not been implemented, I think it deserves being brought up in a .19 feedback thread (considering it might affect grazing).

31.19 does not have animal mass play any factor in how much they eat.  How much they need to eat is derived from an absolute value in the [GRAZER:<integer>] tag.  The integer determines how much of the hunger variable is removed per unit of grass eaten.

I think I recall the hunger variable going from 0 to 65535, with dwarves typically eating at around 45,000 or so, and eating around 10 times a year in a year with 403,200 time units.  I'm guessing this means hunger iterates 1 for every time unit.  Eating a unit of food instantly drops the dwarf to 0 hunger.

Now then, consider that rabbits have 120,000 as their grazer integer.  This means they eat about as rarely as dwarves do, since their hunger shouldn't even be able to go as high as 120,000 in the first place, unless Toady modified something for introducing grazers.

Now consider draltha, a creature with 24 as their grazer integer.  It takes an average of 10 time units just to move from one space to another, and then 10 time units to eat the grass.  They must be in absolutely constant motion, vacuuming up every grass they find, and basically cannot ever stop to skip over a tile that doesn't have grass on it.  Any turn they aren't either eating or moving over another grass tile to eat next turn is a turn they move closer to starvation.

In other words, kill draltha on sight, those things have so much food requirements you can functionally consider them bugged, and will destroy your ability to graze animals.

Plus keep in mind that milk production is also independent of creature size - goats produce just as much milk as cows, in spite of eating less than a tenth as much grass.

I hope that Toady will correct this, and I've already put it in a section of the Improved Farming thread how these things can be fixed.
Logged
Personally, I like [DF] because after climbing the damned learning cliff, I'm too elitist to consider not liking it.
"And no Frankenstein-esque body part stitching?"
"Not yet"

Improved Farming
Class Warfare

Granite26

  • Bay Watcher
    • View Profile
Re: Making more diverse genetics
« Reply #23 on: February 21, 2011, 10:15:34 am »

1:  sterilization has been covered in depth, with lots of excellent research into period techniques.

2:  for rec. genes, why not simply roll 37.5 mom, 37.5 dad, 25% rec., where rec. is rolling again on the initial range chart.  That's a mathematically appropriate approximation.  I might prefer a 10% random item.

3:  a second option would be that the rec trait is rolling probabilities again on the 'loser' parent from the last gen.  Make it recursive, and traits can pop back up at any time, all without tracking any extra information.


NW_Kohaku

  • Bay Watcher
  • [ETHIC:SCIENCE_FOR_FUN: REQUIRED]
    • View Profile
Re: Making more diverse genetics
« Reply #24 on: February 21, 2011, 10:22:41 am »

2:  for rec. genes, why not simply roll 37.5 mom, 37.5 dad, 25% rec., where rec. is rolling again on the initial range chart.  That's a mathematically appropriate approximation.  I might prefer a 10% random item.

3:  a second option would be that the rec trait is rolling probabilities again on the 'loser' parent from the last gen.  Make it recursive, and traits can pop back up at any time, all without tracking any extra information.

Why not just make recessive genes?  They're no more complicated than those two workarounds, anyway.  It just takes a couple more spots in memory and a couple more random rolls than what we have now, and wouldn't involve the random rolls on a full chart or having to trace an ancestry, and would be much more accurate a simulation.
Logged
Personally, I like [DF] because after climbing the damned learning cliff, I'm too elitist to consider not liking it.
"And no Frankenstein-esque body part stitching?"
"Not yet"

Improved Farming
Class Warfare

Granite26

  • Bay Watcher
    • View Profile
Re: Making more diverse genetics
« Reply #25 on: February 21, 2011, 11:01:19 am »

Except we DO trace ancestry and DON'T have recessive genes.  10000 population x 16 byte gene x 30 genes (physical, mental, etc) x 2 copies = 9.6 megs, and thats giving a generously low size per gene.

Even recessive genes aren't sufficient to prevent monoculture in a small gene pool without mutation.  You'll end up with a dominant trait faster, and likely a single recessive trait.

NW_Kohaku

  • Bay Watcher
  • [ETHIC:SCIENCE_FOR_FUN: REQUIRED]
    • View Profile
Re: Making more diverse genetics
« Reply #26 on: February 21, 2011, 11:17:18 am »

Except we DO trace ancestry and DON'T have recessive genes.  10000 population x 16 byte gene x 30 genes (physical, mental, etc) x 2 copies = 9.6 megs, and thats giving a generously low size per gene.

Even recessive genes aren't sufficient to prevent monoculture in a small gene pool without mutation.  You'll end up with a dominant trait faster, and likely a single recessive trait.

It's fine if a very small gene pool is completely taken over by a small number of genes, but it shouldn't be as easy as it is right now.

You're also talking about attributes, which aren't the same things as the hair/skin/eye color "genes" that are in the game right now, which was what I was talking about.  Attributes would need a different system, since you don't just inherit a particular strength value wholesale from one parent or another, anyway.

With that said, you're talking about something that would already be 4.8 megs on just 10,000 population right now.  (I don't think stats are 16 bit, are they?) That seems far too large for the current pre-embark save sizes right now, since I know there is a huge amount of data being spent on just the geology alone. 
Logged
Personally, I like [DF] because after climbing the damned learning cliff, I'm too elitist to consider not liking it.
"And no Frankenstein-esque body part stitching?"
"Not yet"

Improved Farming
Class Warfare

Thundercraft

  • Bay Watcher
    • View Profile
Re: Making more diverse genetics
« Reply #27 on: February 21, 2011, 01:48:06 pm »

31.19 does not have animal mass play any factor in how much they eat.  How much they need to eat is derived from an absolute value in the [GRAZER:<integer>] tag.  The integer determines how much of the hunger variable is removed per unit of grass eaten.
Well, I hope mass gets factored in eventually as that would add much to realism.

Question 1: It has been established that males are capable of breeding via "spores" at a distance with females on the same game map. But are creatures and civs able to mate with each other anywhere on the world map, regardless of how much distance separates them? Or does distance factor in on whether breeding can take place?

Question 2: As of version 0.31.19, is "genetics" limited to the game map in Fortress Mode? Or is it constantly kept track of for all creatures across the entire world map?
Logged

NW_Kohaku

  • Bay Watcher
  • [ETHIC:SCIENCE_FOR_FUN: REQUIRED]
    • View Profile
Re: Making more diverse genetics
« Reply #28 on: February 21, 2011, 01:59:15 pm »

Well, I hope mass gets factored in eventually as that would add much to realism.

Question 1: It has been established that males are capable of breeding via "spores" at a distance with females on the same game map. But are creatures and civs able to mate with each other anywhere on the world map, regardless of how much distance separates them? Or does distance factor in on whether breeding can take place?

Question 2: As of version 0.31.19, is "genetics" limited to the game map in Fortress Mode? Or is it constantly kept track of for all creatures across the entire world map?

I'll break this up between "civilized" and "uncivilized" creatures.

For question 1, civilized creatures only breed through marriage, and only get married to people they share a site with.

Question 2: Genetics regarding the civilization's gene pool is tracked through worldgen, this is why there is so little genetic diversity in civs right now.

Uncivilized 1: As far as I know, creatures that aren't part of civilizations are tracked as a population per biome.  They aren't selective, so as long as there are some males and some females on a site, they would be able to reproduce regardless, but I think each biome is tracked individually, without a chance for cross-breeding between biome populations, although I could be wrong on this.

Uncivilized 2: I think that uncivilized creatures are simply spawned wholecloth when they walk onto your fortress map or you see them, and genetics are not tracked for them as a population, but that when they are on your map, you can do your little eugenics programs.
Logged
Personally, I like [DF] because after climbing the damned learning cliff, I'm too elitist to consider not liking it.
"And no Frankenstein-esque body part stitching?"
"Not yet"

Improved Farming
Class Warfare

Granite26

  • Bay Watcher
    • View Profile
Re: Making more diverse genetics
« Reply #29 on: February 21, 2011, 04:09:09 pm »

Stats are packed crazy tight, but you'd still need a value(color) and dominance(genes are only dominant in comparison to each other).
Pages: 1 [2] 3