Author Topic: Just a quick insight i'd love.  (Read 8128 times)

Offline spike43884

  • Bot Overlord
  • ****
  • Posts: 656
    • View Profile
Just a quick insight i'd love.
« on: May 23, 2015, 12:24:56 PM »
With DB3 im wondering if the following features are going to be implemented?
1)phagocytosis (splitting partially open to ingest something)
2)I've forgotten the specific name for this, but where a cell just pops open, spilling out its guts (not where it splits open to ingest something)
3)Active transport and osmosis
4)Ways to disrupt the phillic and phobic properties in the cell membranes, causing them to, fall apart.
Autism can allow so much joy, and at the same time sadness to be seen. Our world is weird, and full of contradiction everywhere, yet somehow at moments seems to come together, and make near perfect sense.

Offline Botsareus

  • Society makes it all backwards - there is a good reason for that
  • Bot God
  • *****
  • Posts: 4483
    • View Profile
Re: Just a quick insight i'd love.
« Reply #1 on: May 25, 2015, 02:10:56 PM »
I do not see why you will want this, but you can get away with a multibot forming a structure as described and moving smaller bots around. (I think) In general based on what Numsgil is planning the multibotness and cell shapes should be far more interesting.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Re: Just a quick insight i'd love.
« Reply #2 on: May 26, 2015, 12:32:31 PM »
With DB3 im wondering if the following features are going to be implemented?
1)phagocytosis (splitting partially open to ingest something)

Something like you attach yourself to the other bot/shape/thing along a few panels and wait a few turns for a layer to form over the other bot that represents your new membrane.  Then you swap out your new membrane for your old one, and voila, the other bot is now inside you.  A lot of the specifics I won't really know until I start playing with it, though.  It's hard to work out on paper what the best way to do something is.  Usually I need to construct some sort of prototype and play with it.

Quote
2)I've forgotten the specific name for this, but where a cell just pops open, spilling out its guts (not where it splits open to ingest something)

Lysis.  Something like the bot/shape gets decomposed in to some smaller shapes and with nothing to hold them together they'll drift apart.  Again, I won't know a lot of specifics until I play with it some more.

Quote
3)Active transport and osmosis

Something like there are small particles (basically shots) that float around in the fluid and you can sort of "suck" them in through your panels.  You'll be able to filter what sort particles you accept through this method, which is the primary benefit over something like phagocytosis.

Quote
4)Ways to disrupt the phillic and phobic properties in the cell membranes, causing them to, fall apart.

Panels will have something like an integrity rating that gets degraded over time (maybe?) plus from attacks from other bots.  If one of the panels gets too weak it will split and the cell will lyse (either entirely or partially, I'm not sure yet).
« Last Edit: May 26, 2015, 12:34:47 PM by Numsgil »

Offline spike43884

  • Bot Overlord
  • ****
  • Posts: 656
    • View Profile
Re: Just a quick insight i'd love.
« Reply #3 on: May 27, 2015, 05:49:20 AM »
With DB3 im wondering if the following features are going to be implemented?
1)phagocytosis (splitting partially open to ingest something)

Something like you attach yourself to the other bot/shape/thing along a few panels and wait a few turns for a layer to form over the other bot that represents your new membrane.  Then you swap out your new membrane for your old one, and voila, the other bot is now inside you.  A lot of the specifics I won't really know until I start playing with it, though.  It's hard to work out on paper what the best way to do something is.  Usually I need to construct some sort of prototype and play with it.

Quote
2)I've forgotten the specific name for this, but where a cell just pops open, spilling out its guts (not where it splits open to ingest something)

Lysis.  Something like the bot/shape gets decomposed in to some smaller shapes and with nothing to hold them together they'll drift apart.  Again, I won't know a lot of specifics until I play with it some more.

Quote
3)Active transport and osmosis

Something like there are small particles (basically shots) that float around in the fluid and you can sort of "suck" them in through your panels.  You'll be able to filter what sort particles you accept through this method, which is the primary benefit over something like phagocytosis.

Quote
4)Ways to disrupt the phillic and phobic properties in the cell membranes, causing them to, fall apart.

Panels will have something like an integrity rating that gets degraded over time (maybe?) plus from attacks from other bots.  If one of the panels gets too weak it will split and the cell will lyse (either entirely or partially, I'm not sure yet).

The creating new panels is quite a good idea, it'd cover mitosis quite well (in the splitting stage)
With the small floating particles, it'd be nice to have an underlying 'task' for the program to slowly try to reach equilibrium within and outside of the bot. Then of course if the bot wants to create an inbalance, it has to spend some of its energy to perform active transport. It'd be quite nice for some ions in those particles, I'd be tempted to make a multibot which has colonies taking in ions to start electric charges.
Autism can allow so much joy, and at the same time sadness to be seen. Our world is weird, and full of contradiction everywhere, yet somehow at moments seems to come together, and make near perfect sense.

Offline Botsareus

  • Society makes it all backwards - there is a good reason for that
  • Bot God
  • *****
  • Posts: 4483
    • View Profile
Re: Just a quick insight i'd love.
« Reply #4 on: May 27, 2015, 06:15:52 AM »
Off topic: Spike, how do you know biology so well considering your age?

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Re: Just a quick insight i'd love.
« Reply #5 on: May 27, 2015, 12:49:26 PM »
The creating new panels is quite a good idea, it'd cover mitosis quite well (in the splitting stage)

Yeah, my thinking is that a bot starts building panels that split it down some random axis, and when it's ready to properly split these new panels go live and the two daughter bots go on their merry way.

Quote
With the small floating particles, it'd be nice to have an underlying 'task' for the program to slowly try to reach equilibrium within and outside of the bot. Then of course if the bot wants to create an inbalance, it has to spend some of its energy to perform active transport. It'd be quite nice for some ions in those particles, I'd be tempted to make a multibot which has colonies taking in ions to start electric charges.

Ionic transport is really important for cells.  I'd like to have them in the sim, if they served an important purpose, but from what I can tell cells transport ions to "charge" the cell.  The difference in voltage between inside the cell and outside is used to drive chemical reactions.  That is, the ion pumps in a cell are like paying your electric bill: it's not useful in and of itself, but a means to an end (powering your computer so you can play Darwinbots :P).  Plus they'd be somewhat expensive to simulate explicitly.  But I'm open to the idea at least.

Offline Shadowgod2

  • Bot Destroyer
  • ***
  • Posts: 387
    • View Profile
Re: Just a quick insight i'd love.
« Reply #6 on: May 28, 2015, 12:17:14 AM »
Off topic: Spike, how do you know biology so well considering your age?

i knew just as much or more than he does at his age. in fact they teach it in school as early as 5th grade here but most don't care and forget shortly after. only those who really pay attention and/or have a fascination towards biology will retain and build upon it. then again back then i also had 15 plants too.


hay num have you ever seen How to Grow a Planet? there's a few interesting notes you should see if you get the chance especially about chloroplast.
« Last Edit: May 28, 2015, 12:20:26 AM by Shadowgod2 »

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Re: Just a quick insight i'd love.
« Reply #7 on: May 28, 2015, 01:16:49 AM »
hay num have you ever seen How to Grow a Planet? there's a few interesting notes you should see if you get the chance especially about chloroplast.

I've seen it on Netflix but I haven't watched it yet.  I watch a lot of nature programs generally, partly for research and partly because I'm a huge nerd haha.

Offline spike43884

  • Bot Overlord
  • ****
  • Posts: 656
    • View Profile
Re: Just a quick insight i'd love.
« Reply #8 on: May 28, 2015, 05:02:33 AM »
Off topic: Spike, how do you know biology so well considering your age?
Im just generally interested in a lot of subjects really. Plus I have aspergers.

The creating new panels is quite a good idea, it'd cover mitosis quite well (in the splitting stage)

Yeah, my thinking is that a bot starts building panels that split it down some random axis, and when it's ready to properly split these new panels go live and the two daughter bots go on their merry way.

Quote
With the small floating particles, it'd be nice to have an underlying 'task' for the program to slowly try to reach equilibrium within and outside of the bot. Then of course if the bot wants to create an inbalance, it has to spend some of its energy to perform active transport. It'd be quite nice for some ions in those particles, I'd be tempted to make a multibot which has colonies taking in ions to start electric charges.

Ionic transport is really important for cells.  I'd like to have them in the sim, if they served an important purpose, but from what I can tell cells transport ions to "charge" the cell.  The difference in voltage between inside the cell and outside is used to drive chemical reactions.  That is, the ion pumps in a cell are like paying your electric bill: it's not useful in and of itself, but a means to an end (powering your computer so you can play Darwinbots :P).  Plus they'd be somewhat expensive to simulate explicitly.  But I'm open to the idea at least.
Yeah well, you'd primarily need to simulate the attraction to opposite ions first. the further the value is from 0 the more attracted to its opposite it is. Then when ions are formed, energy is used up and when ions are neutralised energy is released, that i'd imagine would be as in depth as it'd be needed. The attraction of ions would be fun to play with.
Autism can allow so much joy, and at the same time sadness to be seen. Our world is weird, and full of contradiction everywhere, yet somehow at moments seems to come together, and make near perfect sense.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Re: Just a quick insight i'd love.
« Reply #9 on: May 28, 2015, 01:33:47 PM »
Then when ions are formed, energy is used up and when ions are neutralised energy is released, that i'd imagine would be as in depth as it'd be needed. The attraction of ions would be fun to play with.

Chemistry/organic chemistry is a little more complicated than that unfortunately.  Many reactions do involve ions (for instance NADP+), but that's actually NADPH being broken down to release energy.  That is, a neutral molecule being broken down (I think that's considered a reducing reaction if I remember my Chemistry, which is doubtful :) ) to ions to release energy.  Chemistry is rather a series of mutations of molecules from one form to another, with the difference in bond energy determining how much energy is required or released.  As far as I know the amount of bond energy a molecule has is not immediately predictable just from the atoms in it (I think it's some complicated function of the various electron shells and you end up getting neck deep in quantum mechanics before long).

Now that said, a simplified chemistry model isn't impossible, but from a computational standpoint I'd want the role of particles that give off a charge to be minimal.  As long as the particles in the sim are passive and only react to the fluid (or electromagnetic forces or whatever), they're O(n) to simulate.  That is, double the number of particles and you double the computation time.  If they also exert push or have mass or generally do anything to influence the simulation at all it starts to move towards O(n^2).  There are ways around that using techniques like multipole methods, but they're really complicated :)

From a gameplay perspective, though, I think you can get a lot of the effects without doing much of the work.  At its core a cell is a capacitor, and uses the voltage difference of the inside and outside of a cell to drive at least some of its reactions.  We could do something like limit how fast things (fats, proteins, etc.) can by synthesized by how large the voltage difference is, which is a function of total charge and bot surface area.  Building up charge could take time.  Expending charge is less limited, but once its gone its gone, and has to be recharged.

Offline spike43884

  • Bot Overlord
  • ****
  • Posts: 656
    • View Profile
Re: Just a quick insight i'd love.
« Reply #10 on: May 29, 2015, 05:31:12 AM »
Then when ions are formed, energy is used up and when ions are neutralised energy is released, that i'd imagine would be as in depth as it'd be needed. The attraction of ions would be fun to play with.

Chemistry/organic chemistry is a little more complicated than that unfortunately.  Many reactions do involve ions (for instance NADP+), but that's actually NADPH being broken down to release energy.  That is, a neutral molecule being broken down (I think that's considered a reducing reaction if I remember my Chemistry, which is doubtful :) ) to ions to release energy.  Chemistry is rather a series of mutations of molecules from one form to another, with the difference in bond energy determining how much energy is required or released.  As far as I know the amount of bond energy a molecule has is not immediately predictable just from the atoms in it (I think it's some complicated function of the various electron shells and you end up getting neck deep in quantum mechanics before long).

Now that said, a simplified chemistry model isn't impossible, but from a computational standpoint I'd want the role of particles that give off a charge to be minimal.  As long as the particles in the sim are passive and only react to the fluid (or electromagnetic forces or whatever), they're O(n) to simulate.  That is, double the number of particles and you double the computation time.  If they also exert push or have mass or generally do anything to influence the simulation at all it starts to move towards O(n^2).  There are ways around that using techniques like multipole methods, but they're really complicated :)

From a gameplay perspective, though, I think you can get a lot of the effects without doing much of the work.  At its core a cell is a capacitor, and uses the voltage difference of the inside and outside of a cell to drive at least some of its reactions.  We could do something like limit how fast things (fats, proteins, etc.) can by synthesized by how large the voltage difference is, which is a function of total charge and bot surface area.  Building up charge could take time.  Expending charge is less limited, but once its gone its gone, and has to be recharged.
Charge we would probably want to remember to make sure as having both negative and positive. we could say inside the cell an 'overrall' charge...A simplified model is of course what we'd have to achieve though.

SLIGHTLY OFFTOPIC: On the original science forums, I saw an interesting topic, someone was questioning why sexual reproduction ONLY have male and female...and it made me think we could actually remove .sexrepro in DB3. because I came up with a system that could generate gametes (with a slight risk of mutation) as far as DB2 goes. Asexually make cells with only half the DNA, the first half or last half...roughly. Then make sure a specific memory location stores which type of gamete it is. Once that is decided when they collide, they inspect that memory location. presuming their compatible, the lets call it 'type one gamete' ties to the 'type two gamete' then pumps most of its energy into the other one, and releases some viruses into its partner that gives its partner all its genes. Boom, complete gene-set made.
Autism can allow so much joy, and at the same time sadness to be seen. Our world is weird, and full of contradiction everywhere, yet somehow at moments seems to come together, and make near perfect sense.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Re: Just a quick insight i'd love.
« Reply #11 on: May 29, 2015, 01:27:09 PM »
SLIGHTLY OFFTOPIC: On the original science forums, I saw an interesting topic, someone was questioning why sexual reproduction ONLY have male and female...

I think some bacteria and yeasts could be described as having more than 2 genders, though only in a very technical sense.  You still only combine two of them at a time to get a new critter.  It's just that there are some different markers they use to prevent mating with close relatives.  It'd be like instead of male/female people only mated with others with compatible blood types.

Quote
and it made me think we could actually remove .sexrepro in DB3. because I came up with a system that could generate gametes (with a slight risk of mutation) as far as DB2 goes. Asexually make cells with only half the DNA, the first half or last half...roughly. Then make sure a specific memory location stores which type of gamete it is. Once that is decided when they collide, they inspect that memory location. presuming their compatible, the lets call it 'type one gamete' ties to the 'type two gamete' then pumps most of its energy into the other one, and releases some viruses into its partner that gives its partner all its genes. Boom, complete gene-set made.

My plan is, since DNA is sorted in to chromosomes anyway (basically logically distinct address spaces, the results of which get merged to run the cell), bots can merge together and pool their resources (and DNA!).  When they go to divide, they can meiosis themselves to split the DNA instead of copying it.  That is, if there are two chromosomes A and a, A gets sent to one bot and a gets sent to another bot.  I use longest common subsequence type algorithms to match up chromosomes based on what their DNA looks like.  When A and a are parting company, they perform a few crossover events, basically the tail of A and the tail of a, at roughly equivalent places where the two match, switch from one chromosome to the next.  You do that a few times and you'be basically shuffled the two chromosomes together.

This goal is to mimic the haploid/diploid generations of yeast, which are far less arbitrary than the strict roles they have in higher organisms.

Offline spike43884

  • Bot Overlord
  • ****
  • Posts: 656
    • View Profile
Re: Just a quick insight i'd love.
« Reply #12 on: May 30, 2015, 05:33:39 AM »
SLIGHTLY OFFTOPIC: On the original science forums, I saw an interesting topic, someone was questioning why sexual reproduction ONLY have male and female...

I think some bacteria and yeasts could be described as having more than 2 genders, though only in a very technical sense.  You still only combine two of them at a time to get a new critter.  It's just that there are some different markers they use to prevent mating with close relatives.  It'd be like instead of male/female people only mated with others with compatible blood types.

Quote
and it made me think we could actually remove .sexrepro in DB3. because I came up with a system that could generate gametes (with a slight risk of mutation) as far as DB2 goes. Asexually make cells with only half the DNA, the first half or last half...roughly. Then make sure a specific memory location stores which type of gamete it is. Once that is decided when they collide, they inspect that memory location. presuming their compatible, the lets call it 'type one gamete' ties to the 'type two gamete' then pumps most of its energy into the other one, and releases some viruses into its partner that gives its partner all its genes. Boom, complete gene-set made.

My plan is, since DNA is sorted in to chromosomes anyway (basically logically distinct address spaces, the results of which get merged to run the cell), bots can merge together and pool their resources (and DNA!).  When they go to divide, they can meiosis themselves to split the DNA instead of copying it.  That is, if there are two chromosomes A and a, A gets sent to one bot and a gets sent to another bot.  I use longest common subsequence type algorithms to match up chromosomes based on what their DNA looks like.  When A and a are parting company, they perform a few crossover events, basically the tail of A and the tail of a, at roughly equivalent places where the two match, switch from one chromosome to the next.  You do that a few times and you'be basically shuffled the two chromosomes together.

This goal is to mimic the haploid/diploid generations of yeast, which are far less arbitrary than the strict roles they have in higher organisms.
Yeah....It'd be nice actually, to see in the DNA viewer the chromosones aligned...so 2 columns, A in one column, then in the same area in the other column a. That and the capability to 'scroll back' through the life of the single organisms DNA, so you can see how its mutated.
Autism can allow so much joy, and at the same time sadness to be seen. Our world is weird, and full of contradiction everywhere, yet somehow at moments seems to come together, and make near perfect sense.