One Hour One Life Forums

</3... ;(

Some things like plant growth rates or anything based on a timed transformation could be influenced
by terrain through a simplified version of how you are calculating temperature mixed with how you
are doing tool use.

If a carrot has say X intermediate growth stages instead of just one (like how tools have use stages
as objects instead of internal object data to track status) then having a chance to progress either by
checking the tile underneath (or if cheap enough/appropriate, tiles in range through a formula like how
temp seems to work) on each transformation cycle, you can give them an effective 'Mean time to Grow'
that varied based on tile/temp etc.

If it is too expensive to run simulation on that many objects you could sidestep it by having the calculation
done when the player initialized the planting eg. If there are X biomes there are X types of seeded watered
carrot plots who have different cycle times, or if there are Y categories of suitability (like {0-5,6-10} ) there
would be Y types of seeded watered carrot plots.

If you have some way of having objects be variations of other objects it would be a lot easier for you to
mess around with balance on your end as a 'desert' well can have it's chance to refill or its cycle time
etc. modded by some variable somewhere on compile. This could also be useful for clothing out of variable
types of material for example.

It is a bummer for me that tiles can't interact directly with each other nor store data (other than through
the crutch of tile types of different names with the otherwise similar properties), but I am still wondering
if tiles themselves can be changed like from grass to prairie etc? Afiak floors and walls are just objects
on top of the terrain and do not change it.

If tiles themselves can be changed through interactions then things like pasturage can be simulated via
something like a sheep turning a grassland to a prairie etc. Even some sort of pseudo inter tile interaction
can be done via intermediaries.

Sheep is hungry, checks it's tile, tile is food, runs feed algorithm on tile, algorithm outputs transformation
to tile type. This could be either single or multi tile and could be initialized by the timed transformations
as well.

My mind is still hung up on animals needing to eat too <3

Failing tiles being changeable, only pasturage simulation I can think of atm is animals eating something
like a 'grass' plant that grows on tiles. I don't know if objects can spawn other objects in other tiles when
they transition on timers or if it would have to be done like berry bushes with picked/unpicked modeling.
A problem seems to be grass being in the way but lower level flora could have an item class like water
containers seem to and have X+LowFlora=X  so you can just drop crap on em if you want the room
(sounds pretty easy to grief is all).

Yeah. Other fun differences might include multiple harvests before needing to be replanted or needing more or less water.

Just as general practice you don't want to install things in "programs" as it is annoying as hell
to get it to stop setting it to read-only over and over and over again. I just put it all in a games
folder I put in C. Windows 10 is rough but sadly so many games need it.

You gotta give someone your email to pm them lol. Sounds like a bad plan...

You could also have the buildings interact with each other:

Perhaps wells near other wells influence each others fill rates negatively.
Too many farms of the same plant in an area could give an increasing risk of blight.
Leaving poop near your food or near your Animals could increase risk of ruining it or them dying to disease.
Wormy rows could spread to adjacent hardened rows over time.
Trees can spawn saplings and saplings can grow into trees.

You can have buildings interact with the environment in both directions:

Wells may fill faster in swamps, slower in deserts perhaps not work at all in the arctic.
Wells cold also influence the areas water table, which in turn can change areas from one biome to another.
Perhaps too much water drained too fast could dry up a swamp or turn grass lands to prairies etc.
Building a pump and an irrigation channel might make the desert bloom.
Higher level techs could melt the arctic and turn ice holes into ponds.
Swampland could water plants very slowly, while maybe farming in deserts could need more water.
Crops can have preferred biomes that influence yields and growth times as well as have the chance to spread to the wild.
Perhaps crops even have biome exclusions so no one village can grow everything, giving a natural reason to trade.
Farming carrots over and over in the same area could influence their growth times, perhaps fertilizer could mitigate this?
Perhaps different types of fertilizer could have different risks, for example poop could spread disease.

Meta justice is theoretically possible. Problem is that there is still no data to weed out murders
that have good cause from those that do not.

There is no real way to bring the accused to testify on their behalf nor is there any real evidence
to examine unless they HAPPEN to have a copy of their game log (before it gets overwritten that is).

Given the game logs, it is theoretically possible for human judges to examine the playbacks (currently
in real time afaik, so super slowly) to see if actions are warranted by either the player or the
community of judges. The current cost to check the work of the judge to ensure their honesty is
equal to the work involved in examining the playback.

This brings you to the problem that you would need even more and more judge work hours to police
effectively. As the number of judges checking each others work increases, the time to judge increases
linearly. 2 judges does not give good results as in cases of ties you either have to assume guilt,
assume innocence or call in more judges.

Then the question is what is our expected population percentages of honest vs dishonest judges?
We need this data in order to decide at what point the returns of more overlapping judges improves
accuracy against the cost of their labor. Even volunteer hours are not unlimited so can not be wasted
endlessly..

Do we decide by majority ruling, unanimous? Do we default to innocent or to guilty?

My expectation is that as judges approaches infinity that the ability to spot good vs bad judges
also approaches 1. You could have a set up a second higher meta level to address the issue of
good/bad judges, but then we are only pushing the problem along even higher as the judges'
judges can also be corrupt.. Granted, one may care less/more about accuracy and settle for algorithmic
justice. Then again if that is the case then it would not be unfair to think this whole enterprise
was pretty useless at that point...

Without any review of murders you could automate it off the data set with ID's, but then even
with that knowledge you could not act in game without being added to the list yourself. The
problem of malicious actors acting as judges fucking with people on this new higher, meta level
is not solved at all. That, and this is ASSUMING you can get access to the game logs of the
Defendant, Plaintiff as WELL as ALL RELEVANT WITNESSES... The data set of these people could
theoretically be trawled from the family tree within +/- 1 hour from time of death, but that would
not work well in multi family cities as far as the witnesses go.

There is no way currently for a computer program to automatically determine whether someone is
guilty or innocent when judging inside of a shifting contextual environment. One of the limits
of computation is that it is inherently NOT self aware and is thus unable to perceive context
or meaning at all. You can in some instances give it parameters to make judgement inside of
but only if the context is fixed enough to be able to pre-program it in.

Given all this a black list of murderers only pushes the meta problem of murder/grief/justice/trolling
up one level higher without easing our ability to solve it.

Justice is an interesting problem/concept in both the real and theoretical application of it.

Imagine the case where there are multiple instances of overlapping blacklist based justice systems
cross blacklisting each others user-bases to achieve dominance to promote their own conceptions
of justice on yet a higher meta level..

The needless complexity is just BEAUTIFUL.... <3

She gonna get fired...

Not gonna work so well unless you are gonna either sing it or give sheet music etc.

Also good thread. If you only copy good threads you will only make good threads..

Not a bad strategy <3

                    else if( ourLiveObject->foodStore <= 3 &&
                             computeCurrentAge( ourLiveObject ) < 57 ) {
                        
                        // don't play hunger sounds at end of life
                        // because it interrupts our end-of-life song
                        // currently it's 2:37 long

https://github.com/jasonrohrer/OneLife/ … fePage.cpp

No thanks.
Inevitable doom isn't fun, there is no point investing in the future if it is predetermined, either good or bad.

What is that? Does it like automap the terrain?

Hell yeah

This was cool story, thank you!