Darwinbots Forum

Bots and Simulations => Evolution and Internet Sharing Sims => Internet Mode Commentary => Topic started by: MacadamiaNuts on December 14, 2007, 09:55:23 AM

Title: The State of the Simulation
Post by: MacadamiaNuts on December 14, 2007, 09:55:23 AM
Open topic for updates about what's going on in the IM sims.

Multiply is ruling with Lionfish 5.x subversions going up and down. And stickerbush is still the king of the green hill.
Title: The State of the Simulation
Post by: Peter on December 14, 2007, 10:33:36 AM
Nice idea, I suggest too that everyone post some graphs of is happening in IM, I like graphs. I can't post graphs as I still can't connect, I'd like to see what's happening.
Have commentaries of IM on other topice and keep this one just a overview fromwhat is happening in IM, keeps it cleaner.

Quote from: MacadamiaNuts
Multiply is ruling with Lionfish 5.x subversions going up and down. And stickerbush is still the king of the green hill.
Wow, really multiply, then someone probably introduced it. Happy to see it is surviving, and evolving could very well get better bots, as it is far from optimal, as I just optimised it by deleting or adding zeros  , talking about effiency.
Title: The State of the Simulation
Post by: EricL on December 14, 2007, 11:04:04 AM
Windows rebooted EricL3 in the night and the sim didn't get saved in time, so I had to take that machine back in time a day or two which is where all the new Flypaper 3.2's in IM just came from...
Title: The State of the Simulation
Post by: MacadamiaNuts on December 14, 2007, 12:42:27 PM
Now its:

748 - Lionfish 5
501 - stickerbush
379 - multiply
276 - Flypaper 3.2
167 - plant
118 - Alga_Minimalis
88 - Lionfish 5.2
78 - multiply 5199
44 - Lionfish 6
43 - algadillo
38 - anemone
35 - Lionfish 5 9149
14 - multiply 1525

Lionfish 5 seems to be in EricL2 and EricL3. Yesterday I found that the body feeding and the self evolution were seriously broken, so I updated it to 6.

LF5 breaks the birth tie too early, so it always defaults back to vanilla then mutates once. Also only big bots were using the body feed. In LF5.2 I added learning. They broadcast their evolutions one at once, and younger bots copy them.
Title: The State of the Simulation
Post by: MacadamiaNuts on December 14, 2007, 01:30:41 PM
Right now Flypaper 3.2 is on a spike up to 3800.
Title: The State of the Simulation
Post by: MacadamiaNuts on December 15, 2007, 10:07:39 PM
I lost yesterday's graph.  

Now there's:

3018 - Flypaper 3.2 (that sim has been stuck since yesterday)
976 - multiply (it has been spiking around 1500 all the time)
883 - stickerbush
530 - Alga_Minimalis
458 - Lionfish 5 (it entered my LF6 sim and now owns half of it   )
271 - Lionfish 6
And some under 11 that are in the stuck sim too.

I had recently teleported in:
multiply from EricL2 and EricL3.
Lionfish 5 from Bot House and EricL3, and it's in my sim now too.
Lionfish 6 from EricL. There are over a hundred not in my sim.
stickerbush from everywhere.  
Title: The State of the Simulation
Post by: Peter on December 16, 2007, 04:53:35 AM
If the pop files are stuck, a simple fix is to delete them, they create themself again after. Atleast that worked in earlier versions.
Title: The State of the Simulation
Post by: EricL on December 16, 2007, 02:59:43 PM
LF6 seems not to have taken hold and appears extinct in the currently connected sims.

Melangis(Garet) currently at the top of the heap though that species is concentrated in a single newly connected sim (presenty experiencing a population spike) and has yet to expereince much interaction with other species.  Then it goes:

multiply
Shrinking Violet
stickerbush
LF5
Alaga
and some others in the double digits

I introduced Shrinking Violet last night.  Very simple (66 bp) one gene organism (used new conditional paradym) but with some interesting implications.  It's totally defensive.  Flees from everything, super small for speed and to avoid being shot, plays by different rules than stickerbush and should displace it.  Uses omnieye targetting, so may not rule large sims until 2.43w.  But I expect it will become the dominate veggy.  It's very hard to catch and should have serious implications on up the food chain.  

I'm telling you, veggies are too powerful...
Title: The State of the Simulation
Post by: Garat on December 16, 2007, 03:41:09 PM
Yes, Melangis has to experience much interactions.


I'm not sure that he is somewhere else than on my computer
Title: The State of the Simulation
Post by: Peter on December 17, 2007, 04:39:50 PM
Current standing

Melangis(Garai) ~2000
Multiply(peter)   ~1800
Shrinking Violet ~700
Lionfish 5          ~450
Lionfish 6          ~450
Alga minimalis   ~300
Lionfish S-V       ~260
Beholder           ~250
Stickerbush       ~100
Multiply(simple) ~100
Title: The State of the Simulation
Post by: shvarz on December 17, 2007, 08:07:54 PM
I had a huge spike of Melangis, but then lionfish 6 came in, killed off the whole sim and died   I kind of liked these little bastards...

Now I have mostly multiply and some Lionfishes. But Lionfish population is fairly stable, so I'm hoping to get a complex ecosystem going. See the attached picture. (multiply is green, alga is brown and lionfish are blue)
Title: The State of the Simulation
Post by: EricL on December 17, 2007, 08:32:26 PM
multiply and LF seems to co-exist/co-evolve pretty well together.  LF can take on a bunch of multiply's, but too many and LF gets overwhelmed.  Neither seems to be able to wipe the other out completley in my sims.  

But I'm wonderring how they will do longer term without a ready source of easy nrg from fat, stupid veggies.  Shrinking Violet has largly displaced stickerbush.  There's a bunch of other veggy species from new or periodicly connected sims, but I predict they won't last.  It's just a matter of time before SV displaces them, particularly once 2.43w comes out (a day or two) and omnieye navigation works on larger fields.  In EricL for example, which is running a pre-release 2.43w, the population has fallen to the point where the cost multiplier is 0.  Multiply and LF survive in low numbers, preying on each other and the occasional teleported in veggy or mutated SV beg bertha.  Catching a live SV is rare and generaly a function of luck, trapping it against a shape for example (still have to fix wide-eye shape recognition).

My goal with SV is to raise the stakes on what it takes for a bot to be successful in IM.  No more fat, easy veggies lying around providing easy nrg for relatively stupid bots.  People wanted full function veggies, you got em.  Now to be sucessful, your bot has to either prey on non-veggies (which themselves have to exist in soem numbers and get their nrg from somewhere) or catch SV's.  Someone is goign to have to catch SV's - no easy task.  I suggest strategies such as herding cooperation between multiple hunting bots, ambushes from behind shapes, corning in the corners of non-tooridal sims and so on.  Good luck.

I told you, veggies are too powerful...
Title: The State of the Simulation
Post by: Numsgil on December 17, 2007, 09:25:40 PM
Quote from: EricL
I told you, veggies are too powerful...

Heh, I like that post script at the end of all your posts.  I say, bring it!  If and when it is shown that a veggy exists that is simply broken and cannot be defeated, we can make some changes.

I am extremely tempted to devote my pre-Christmas down time to constructing a novel bot strategy.  Maybe some sort of phalanx to mow down those barbarians.  The DNA is increasingly getting to the ideal I imagined when I first joined.  If only the physics were as easy
Title: The State of the Simulation
Post by: EricL on December 17, 2007, 09:44:18 PM
Oh, no need to spend yoru vacation that way.  There's a real easy way to defeat SV.  Just turn off the sun! Veggy repop may (eventually) supplant it in your sim.  But of course, if everybody turns off the sun, then that effectivly illiminates the advantage of being a veg in IM, which is what I wanted in the first place!!!!  Ha hah!!!!!  Can't you see the beauty, the billiance of those simple little 66 base pairs???!!!!  I feel so.... diabolical!!!

Did I mention that veggies are too powerful?
Title: The State of the Simulation
Post by: shvarz on December 18, 2007, 12:11:19 AM
Actually, giving energy proportionally to the body and giving only energy (not body) handicaps SV severely. It came to my sim, but could not displace "fat stupid Alga Shellular"
Title: The State of the Simulation
Post by: MacadamiaNuts on December 18, 2007, 08:09:12 AM
Thick fluid physics are funnier. ^^
Title: The State of the Simulation
Post by: MacadamiaNuts on December 18, 2007, 09:51:34 AM
[attachment=764:attachment]

Lionfish 6b wishes you merry end of 2007.
Title: The State of the Simulation
Post by: shvarz on December 19, 2007, 01:47:20 PM
I'm having a huge spike of anemones right now. I expect the population will crash fairly soon. I also have viruses spreading like crazy.

I have a bit of Lionfish around, multiply is gone. Just a single Shrinking Violet and it barely survives - most of my veggies are Alga Shellular. Anyone else have these? They are very inefficient and defenseless and take forever to grow and divide, so I don't imagine they survive in other people's sims.
Title: The State of the Simulation
Post by: Peter on December 19, 2007, 02:25:47 PM
Quote from: shvarz
I'm having a huge spike of anemones right now. I expect the population will crash fairly soon. I also have viruses spreading like crazy.

I have a bit of Lionfish around, multiply is gone. Just a single Shrinking Violet and it barely survives - most of my veggies are Alga Shellular. Anyone else have these? They are very inefficient and defenseless and take forever to grow and divide, so I don't imagine they survive in other people's sims.
The grow in virus has probably coused the lower population of multiply. I haven't got a serious anti-virus-system in them, they do have a passive no-repro defence, but I have noticed that's the first thing that goes broke in evosims.

And about the veggies, I had them  . Mainly becouse it was the first bot that came in  , But I have had them.
Title: The State of the Simulation
Post by: MacadamiaNuts on December 20, 2007, 03:59:25 PM
1522 - multiply(peter)
593 - Shrinking Violet
488 - Lionfish 1.4 (!?)
376 - anemone
219 - Alga Shellular
120 - anemone 1256
116 - Lionfish 6
66 - Lionfish 5
20 - Explosive shrinking violet
10 - Mutatis
etc.

I've been puzzled all day about LF 1.4. That's a quite old bot, I could swear I haven't loaded it in the last month. Who's running it?

EDIT: Ok, I saw one coming from EricL2. Yet, it must have entered there from somewhere else. :?
Title: The State of the Simulation
Post by: Peter on December 20, 2007, 04:04:11 PM
Quote from: MacadamiaNuts
I've been puzzled all day about LF 1.4. That's a quite old bot, I could swear I haven't loaded it in the last month. Who's running it?
Well it is beating your lionfish in numbers  

Same thing about multiply, it is old(well realtive) and I did't introduce it last month.
Title: The State of the Simulation
Post by: MacadamiaNuts on December 20, 2007, 04:52:06 PM
That's sort of a General Rule of DB:

The most complex you do a bot, the duller it is at surviving.
Title: The State of the Simulation
Post by: EricL on December 20, 2007, 06:05:47 PM
A bunch of old bots  - LF 1.4, Pred9 - came in from Num's machine.  He hadn't connected to IM in a long time and the last time he did, I still had a bug where I would teleport out more than would be uploaded, so bots would build up in the local out directory for fast running sims where the cycles per upload was high.  He probably had a bunch  - maybe hundreds - still there that got uploaded.
Title: The State of the Simulation
Post by: Numsgil on December 20, 2007, 07:03:17 PM
My Darwinbots install has problems with the internet connection.  After a few hours the FTP control breaks until I reset the program.  So a lot of the time my sim runs unconnected for a while
Title: The State of the Simulation
Post by: MacadamiaNuts on December 20, 2007, 07:22:14 PM
That happens to me too, I thought it was only my punishment for turning into a linux weirdo.

The above thing is why I like local in/out teleporters. You can store there all sorts of bots and then get them back mixed together.

Life is like a DarwinBot's inbound 'porter --you never know what you gonna get.
Title: The State of the Simulation
Post by: Peter on December 21, 2007, 04:12:17 PM
Current standings

1965 - Duplo
1780 - Lionfish 1.4
809 - Lionfish 4
607 - Shrinking Violet
248 - Alga_minimalis
61 - Mutatis
34 - Lionfish 5
22 - Multiply(peter)
16 - Ex. Shrinking Violet
6 - Preditor9(EricL)
2 - anemone

Multiply has fallen low, if I look now there are only 22 left. Could someone grap the dna of it, I'd like to see what have changed. Change isn't very big I'll see it teleporting in in my sim.
Title: The State of the Simulation
Post by: MacadamiaNuts on January 28, 2008, 10:56:08 AM
Mmh... the population files I receive are always the same. I tried to delete them manually and wait them to be downloaded again, but got the same stats.
Title: The State of the Simulation
Post by: EricL on January 28, 2008, 11:08:14 AM
I wait between 24 and 48 hours before deleting server-side pop files for inactive sims.  When is a sim really gone as opposed to just down for a bit?  24 hours minimum, 48 max was a simple easy choice.

Try it now.  I deleted them by hand.
Title: The State of the Simulation
Post by: Moonfisher on February 21, 2008, 10:28:54 AM
I still can't get internet mode to work...
I switch on the F1 internet mode flag, then the log sais it's deleting the old pop files, then once the first countdown is over I get the graph.
But after that... nothing happens, the only values on the graph to change are the ones for bots that I'm running myself.

I wanted to see if Pacifist could handle the 25K internet challenge, but if I never send out my bots I don't see how that would happen.
Tryed just making a max size environment with 300 alge, but everything just froze up completely around 9K while the bot was still just getting started...
I suspect that my bot may have problems spreading through teleporters (Since they're always small and average 0-250 nrg while hibernating below 100).
I guess you could do all sorts of stuff, like disable hibernation below a certain myspecies count, or save up energy above a certain myspecies count, but it seems "unethical" to try to adjust the bot to be well suited for teleportation...
I'm hoping the infected alge minimalis would be able to clear the way, or atleast survive so it can feed incoming pacifists...

Either way it doesn't matter if I can't get internet mode to work...

Would be cool if someone would try out Pacifist on internet mode and tell me if it has a chance...
Title: The State of the Simulation
Post by: rsucoop on February 21, 2008, 01:55:14 PM
Quote from: Moonfisher
I still can't get internet mode to work...
I switch on the F1 internet mode flag, then the log sais it's deleting the old pop files, then once the first countdown is over I get the graph.
But after that... nothing happens, the only values on the graph to change are the ones for bots that I'm running myself.

I wanted to see if Pacifist could handle the 25K internet challenge, but if I never send out my bots I don't see how that would happen.
Tryed just making a max size environment with 300 alge, but everything just froze up completely around 9K while the bot was still just getting started...
I suspect that my bot may have problems spreading through teleporters (Since they're always small and average 0-250 nrg while hibernating below 100).
I guess you could do all sorts of stuff, like disable hibernation below a certain myspecies count, or save up energy above a certain myspecies count, but it seems "unethical" to try to adjust the bot to be well suited for teleportation...
I'm hoping the infected alge minimalis would be able to clear the way, or atleast survive so it can feed incoming pacifists...

Either way it doesn't matter if I can't get internet mode to work...

Would be cool if someone would try out Pacifist on internet mode and tell me if it has a chance...

I have ahd the same problem before. Its just that the computer and the server can't establish a secure enough connection; could be a weak line or an overcrowded server. When it works you get a teleporter, I've had little success because their are no real strong anti-viral genes out there right now, so the Lionfish comes in and destroys entire sims. I would suggest making a stronger antiviral gene before using your Pacifist.
Title: The State of the Simulation
Post by: Numsgil on February 21, 2008, 08:34:34 PM
We've been having technical difficulties with the program and server.  We aren't totally sure if it's the server of the program, so it's been hard to fix.
Title: The State of the Simulation
Post by: rsucoop on February 21, 2008, 09:15:03 PM
Quote from: Numsgil
We've been having technical difficulties with the program and server.  We aren't totally sure if it's the server of the program, so it's been hard to fix.

Have you tried ping requesting the server? It could be that the weather has interupted whatever network its on.
Title: The State of the Simulation
Post by: Numsgil on February 21, 2008, 09:53:59 PM
web server (forum) and ftp server are run from the same place, so if one were down they'd all be down.  We're thinking either it's a problem with the Visual Basic control, or with the number of sessions from the same IP.
Title: The State of the Simulation
Post by: rsucoop on February 23, 2008, 02:59:55 PM
Quote from: Numsgil
web server (forum) and ftp server are run from the same place, so if one were down they'd all be down.  We're thinking either it's a problem with the Visual Basic control, or with the number of sessions from the same IP.
Is the FTP on a seperate switch or router, or is it all the same?
Title: The State of the Simulation
Post by: Numsgil on February 23, 2008, 09:25:01 PM
All the same, as far as I know.  The FTP for the program uses an FTP account that gets limited access to a folder on the web server.  However, GoDaddy (our host) has a two session limit per IP thing.  So at least part of what I think happens is that Darwinbots loses track of a session, and after  a while tries to create a third session and fails.

That doesn't explain why some people can't get even a first session going, of course, so I'm a little stumped.  What I need to do is create a test app in C# that uploads and downloads to/from the server, to check if it's something bogus with the VB controls, or with our server.
Title: The State of the Simulation
Post by: rsucoop on February 25, 2008, 10:01:04 PM
Quote from: Numsgil
All the same, as far as I know.  The FTP for the program uses an FTP account that gets limited access to a folder on the web server.  However, GoDaddy (our host) has a two session limit per IP thing.  So at least part of what I think happens is that Darwinbots loses track of a session, and after  a while tries to create a third session and fails.

That doesn't explain why some people can't get even a first session going, of course, so I'm a little stumped.  What I need to do is create a test app in C# that uploads and downloads to/from the server, to check if it's something bogus with the VB controls, or with our server.

I've noticed that when I didn't shutdown properly from DB when in IT mode, it did not always start up again on restart. My guess is that the VB controls aren't reseting the IT settings when this happens, so the server thinks its still there. Also, the server could be confusing the portion of the bot called, Incoming, where it can say RSUCOOP after its gone through teleporter a couple of times. So you end up with all these identical IPs on bots all over the place, I'm guessing it could be confusing references with the bots, casuing duplicating IPs. If it was still there after you loged off, it could be reading the server as full.
Title: The State of the Simulation
Post by: Moonfisher on February 26, 2008, 02:49:24 AM
Quote from: rsucoop
I would suggest making a stronger antiviral gene before using your Pacifist.

Pacifist uses no antiviral gene or slime, it doesn't need it.
It's problems in IM aren't the viruses, it's more the fact that it's a pacifist, so if an enemy is present and controlls the alge it won't be able to get to the food.
Works well in leagues though, except it takes a while, again because it's a pacifist

And it sounds like I'm having the whole "not shutting down correctly" problem... DB always freezes when I try to close it after IM.
Title: The State of the Simulation
Post by: EricL on April 04, 2008, 01:04:26 AM
A_Supremus has taken over IM.  It managed to wiped out Preditor13, Locust and Guardian in short order today and now owns both my sims.  

On the veggy front, the non-mutating Shrinking Violet 2 has reined supreme for weeks now, but was never able to completly exterminate the orginal, mutating version Shrinking Violet.  It's popualtion was below 100 for quite a while, but it has now made a convincing comeback, largly taking over the EricL2 sim and making good gains in the much larger EricL sim.

The perf improvements in 2.43.1g have allowed me to double the population target for EricL to 1500 and it still runs considerably faster than it did in previous versions.
Title: The State of the Simulation
Post by: asterixx on April 04, 2008, 03:23:26 PM
I've been trying a Zerobot simulation so I've neglected to use the IM over the past few days. It doesnt surprise me that the Shrinking Violet is still dominating, it really has a way of taking over.
Title: The State of the Simulation
Post by: Moonfisher on April 04, 2008, 06:04:33 PM
Shrinking violet is anoying  damn slippery !
When I started working on Tiefighters I had shrinking violet in mind, but the tie angles on multibots got messed up when trying to tie, so had to use reproduction to spawn 2 bots and make 3 ties at the same time, one to prevent the tie prom breaking, and 2 to moake the alge feed you all the energy.
But appart from the thick slimelayer keeping ties off, the repro thing was just way too slow to hit shrinking violet. (Still looks cool when it works though, like a deathstar beam or something)
Could be fun to try to make an alge, they seem like they would be overpowered, but the veggy cap can be realy inhibitting. If you make a combat veggy it just risks loosing a bot leaving SV a chance to reproduce and fill the slot. I might try toying with the idea of attacking anything that flees from us and flee from anything that attacks us, maybe test ways to reconize dangerous bots, or something like that, seems to work well in nature.
Had an algy virus at one point that simply set it's own aim to enemy's aim and used same dx value and some extra to up... would make it chace anything that flees and flee from anything that attacks. But it was more effective to just go one way in that specific case.
Either way I think maybe a system like that could be effective to test oponents to see if we're prey or predator.
Obvisouly that system can be exploited, anyone else see that video with the chubby white guy who crawls up to some lions with a toilet roll as his only weapon ? The lion just doesn't know what to do... he looks wierd and acts wierd... not scared... not agressive... just wierd and relaxed... and he has a toilet roll... if it was human it would still be confused !
Title: The State of the Simulation
Post by: EricL on May 01, 2008, 03:59:02 PM
Seasnake 1.2 and Spinner 1.3 continue to duke it out as they have for days.  Seasnake owns one of my sims, Spinner the other.  Neither seems to be able to teleport enough soliders into the other sim to take it over.    I've seen populations of the minority bot reach 50 or so in either sim before being beaten back by the owner.  Duplo remains a periodic entry, long surviving in Shvarz's periodicly connected sim.  It seems to be suited to whatever the conditions there are, appearing in other sims rarely and not long surviving there but able to defend it's own territory effectively.

Other bots show up in small numbers occasionally but those sims don't stay long and those bots have been unable to colinize any of the permanent IM sims.
Title: The State of the Simulation
Post by: shvarz on May 02, 2008, 01:38:41 PM
Eric, I'm getting strange Alga Minimalis from you - some of them have a huge single gene and others have no DNA at all. Weird.
Duplo is doing quite well in my sim, the "follow the leader" behavior that I described elsewhere is still present. But it figures that it does not survive in other sims - it's adapted to conditions with no costs other than body upkeep and movement.
Title: The State of the Simulation
Post by: EricL on May 02, 2008, 03:36:19 PM
Quote from: shvarz
Eric, I'm getting strange Alga Minimalis from you - some of them have a huge single gene and others have no DNA at all. Weird.

That big single gene is the Seasnake 1.2 genome.  Your getting Seasnake infected veggies as well as ones that have been infected, subsequently mutated chanign their dna length and then committed suicide (by deleting their single gene).
Title: The State of the Simulation
Post by: Trooper5445 on May 06, 2008, 10:55:05 PM
That seasnake is evil. Brought my sim (and soon my computer) to its knees. Anyhow can we get an update. I sent the latest build of my L_Sesillia on it and last I checked it had a small stable community but it might have been eliminated since then.
Title: The State of the Simulation
Post by: EricL on May 07, 2008, 01:35:26 AM
EricL2 continues to be ruled by Spinner V1.3.  The snake makes occasional inroads, but Spinner beats it back.

EricL where the snake resides has been up and down the past few days as I work on a really nasty bug having to do with locked arrays and a bug in VB where if you exit or goto out of a With block, the variable or array or whatever referrred to in the With statement remains locked and cannot be redimensioned.  A real royal pain this one is as I work to make the rob array dynamic and thus remove the 10k bot limit in sims (as well as shrink the working set considerably for smaller sims).  Think I have it now though.  2.43.1k out tomorrow hopefully.
Title: The State of the Simulation
Post by: Numsgil on May 07, 2008, 02:35:40 AM
I remember that bug.  Ugh.
Title: The State of the Simulation
Post by: Trooper5445 on May 07, 2008, 11:31:23 PM
Ah the simulation is interesting. Bloody seasnake. Who created that monstrosity? Anyhow my bot is doing rather well. About as well as spinner. Of course every time L_Sesillia gets ready to increase in pop the snake kills too many of my bots. Oh well its an existence.

How does it look? All I have is a population graph to guess at. I'm going to switch L_Sesilli to Purple as well for easier recognition.
Title: The State of the Simulation
Post by: EricL on May 08, 2008, 10:06:52 AM
Seasnake 1.2 finally managed to colonize EricL2 and wipe out Spinner 1.3.  Not bad for a multibot...

Respect the snake.
Title: The State of the Simulation
Post by: shvarz on May 08, 2008, 01:07:24 PM
Ever since I got those weird "Alga Minimalis infected with Seasnake genome" veggies, I had only them. They did not necessarily wipe out the rest of the bots, but they just outcompeted for energy my original "alga shellular" and "Duplo" died out because it was adapted to eating "Alga shellular".  These weird veggies are mutating like crazy, which might suggest that either a) they don't need most of the functions in their genome for survival or B) they are adapting to the new environment. They also learned to extract energy from nowhere grew to huge numbers in my sim, but after Eric fixed the waste bug, they are back at ~600 individuals.

Stuff keeps coming in from other people, but nothing survives for long.
Title: The State of the Simulation
Post by: shvarz on May 11, 2008, 05:01:55 PM
I was looking at the "infected Alga minimalis" and noticed a specific phenotype - they group into twos, come on top of one another and stay that way turning around and around, constantly shooting stuff at each other (I think it's waste). See attached picture.  

I am curious - is this a normal behavior for veggies infected with Seasnake DNA?  Or did it develop specifically in my sim?  Anyone else see this happening?

BTW, Nanite Detonators seem to be doing OK in my sim, they are not growing fast, but at least they are not dying...  They are managing to get by...
Title: The State of the Simulation
Post by: Trafalgar on May 11, 2008, 05:12:46 PM
Quote from: shvarz
BTW, Nanite Detonators seem to be doing OK in my sim, they are not growing fast, but at least they are not dying...  They are managing to get by...

My 2.2 version wipes out everything I put it against in my sim, but as far as I can tell from the internet population thing, they aren't doing well anywhere else.

I even started a sim up with nothing but seasnake 1.2 and algae minimalis, waited until the seasnakes were dominant and swimming around with 6 or so bots per snake, and then introduced a few nanite detonators, which proceeded to spread and eventually take over the sim (There are about 67 seasnake bots left, in the upper-left and lower-left corner of the sim. There were 80-something the last time the graph updated before that).

What conditions are y'all using in your sims? (I'm using standard F1 conditions, I think)

Edit: Though after they reach 1000 bots of the same species in the sim, they mostly just sit there and do nothing. (Putting most of their energy into body)
Title: The State of the Simulation
Post by: Numsgil on May 11, 2008, 05:15:42 PM
shvarz uses some tweaked cost values.  I think it's just cost for body, IIRC.
Title: The State of the Simulation
Post by: EricL on May 11, 2008, 05:25:24 PM
Both EricL and EricL2 are running pretty standard physics with a few visible shapes.  Bots that arn't shape aware tend to bash themselves to death.

Neither sim has sunlight.  Veggy repop is the only source of nrg.  This keeps the seasnake genome and other virus infected veggies from taking over.  Also limits the utlitiy of battery bots.

The only cost is an age cost but auto-costs tend to keep it high, like 1-2 nrg per cycle.  Seasnakes are good in high cost environments since they store a lot of nrg in body and rachet this down over time.  When a snake dies from running out of nrg, the whole snake dies all at once.  The cells share nrg until then.

Nanites show up and get to maybe 100 or so, but the high cost environment tends to keep their numbers down.  High age costs work against cancerous and semi-cancerous bots.   They favor more robust, higher nrg, higher body macro organisms.  The more bots you use to spread your nrg around, the faster the costs deplete it....
Title: The State of the Simulation
Post by: Testlund on May 11, 2008, 06:36:13 PM
Quote from: EricL
Seasnakes are good in high cost environments...

Not in my sim.    Everything that comes in dies quickly except for Alga Minimalis that now slowly increase my population. I'm running with morphological costs only. No upkeep or ageing costs. Costx has been staying at 26.83383 fro the last days.
Title: The State of the Simulation
Post by: EricL on May 11, 2008, 11:47:58 PM
Quote from: Testlund
Not in my sim.    Everything that comes in dies quickly except for Alga Minimalis that now slowly increase my population. I'm running with morphological costs only.
Well, sure.  If all you want is pond scum, using morphological costs is a fine way to select against complexity.
Title: The State of the Simulation
Post by: Testlund on May 12, 2008, 12:29:00 AM
You all seem to run combat bots designed to work without costs while I run evobots, so it's quite different.
But logically complexity should be able to evolve with morphological costs. It's just that the bots need to do it in the right order. After all, swimming and running isn't free in the real world. Anything living need a way to gain energy for it. All it should take here is a bot to be able to fire a tie and do tie feeding to get the extra energy to move around more, or shoot -1 shots when it bumps into a bot. For some reason info shots are more common than feeding shots. What choses to come out of mutations are maybe not random. Never seen a zerobot evolve tie feeding for instance, though I have had a whole network of zerobots tied together in my sim last year.
I'm hoping though that complex behavior will evolve from very simple in as natural environment as possible.
Also my kind of bots have to have costs otherwise they'll freeze my sim with too many bots.
Title: The State of the Simulation
Post by: Trafalgar on May 12, 2008, 01:27:06 AM
Quote from: EricL
Both EricL and EricL2 are running pretty standard physics with a few visible shapes.  Bots that arn't shape aware tend to bash themselves to death.

How DO you get shape-aware bots, anyhow? I've spawned some random shapes, and as far as I can tell, the bots can run right into them and still not see them - reftype is 0 still and the vision distance isn't appearing shorter like it does when it sees a bot.

Though as far as I can tell, running into the shapes isn't doing any damage either.
Title: The State of the Simulation
Post by: EricL on May 12, 2008, 01:39:18 AM
I need some new shape UI.  For now, checking the Bots Can See Shapes checkbox on the New Shape UI will make all shapes visible.
Title: The State of the Simulation
Post by: Testlund on May 12, 2008, 12:17:53 PM
Too many Alga Minimalis have started to fill my sim, so I decided to set costs back to 0 and added ageing costs. Maybe some of your bots will come in and eat the algas before population reaches 3200 where costX will start to increase.  
Title: The State of the Simulation
Post by: Moonfisher on May 12, 2008, 05:26:34 PM
Heh I saw spinner v1.3 was down to 1 bot, and then fought it's way back up to 30 bots.
It's the first bot I've had any success with on IM, so figured I'd try to see if those teleporters where working for me so I could send some reinforcements.
And it seem like I have no problems sending and recieving bots in version k.
So started 2 sim with a veggy population of 2 and spinner v1.5.

Right now I have somewhere around 500-600 bots in my 2 sims, and theres about 2500 spinners out there, and seasnake is down to about 300 (It was almost at 3000 when I started the sims)
It took about 20000 cycles  
Finaly got to successfully spread something on IM !  
Title: The State of the Simulation
Post by: EricL on May 12, 2008, 07:12:15 PM
Yup.  Spinner 1.5 has made the snake extinct in EricL2 and it's on it's way out in EricL.  Nicely done.

The snake is dead.  Long live the snake!
Title: The State of the Simulation
Post by: Moonfisher on May 13, 2008, 01:03:11 PM
It looks like Spinner never fully eradicated Seasnake 1.2, I still see about 100 of them left.
It also looks like Seasnake 1.3 is advancing slowly but surely, it was at 30 when I got on and it's over 100 now...
It's not certain that it will push back spinner, but it looks like that's what happening.
So it looks like spinners reign won't last much longer, was nice while it lasted
Title: The State of the Simulation
Post by: Testlund on May 13, 2008, 01:47:00 PM
Sorry for getting longwinded about my costs but Sea Snake 1.2 seems to have evolved to handle a small amount of my morphological costs. It was swimming
around and feeding for awhile with CostX at 1.12024. Then later they were all gone. I think the reason for that is my bots are spread out a lot, so they couldn't keep getting enough food in the long run.
They don't appear to be snakes any longer though, more like single bots that follow each other and on occation form ties to it. Also they tend to shoot info shots into alga causing the alga to swim away from them. Not so smart I think.
I just came back to check on my sim this evening and I have several colonies of Alga Antiviral here and there.
Now my costX has started to increase again (4.9 and counting), so only conservative bots will survive now.
You may wonder why I'm not getting bored with such a sim, but I find it kind of exciting to see what ever evolves. It doesn't have to be aggressive hunters for me to like it, but it would be nice if I could at least get a tie feeder. Sigh.
There was this sleep scientist once that said that "the meaning of life is to sleep. There's no point in running around and burn a lot of energy". That's how all life more or less behave on the planet; not to burn more energy than absolutely needed, because life is costly. Waste too much energy and you die. The more costly the environment the more conservative you need to be. That's why the Grizzly bear sleeps half the year, and the other half it constantly eats so it can sleep again. Sleeping, eating and reproducing is what life is about.  
Title: The State of the Simulation
Post by: Numsgil on May 13, 2008, 02:14:09 PM
My three favorite things
Title: The State of the Simulation
Post by: Testlund on May 13, 2008, 02:52:11 PM
I thought you programmers didn't have time with any of that.  
Title: The State of the Simulation
Post by: Numsgil on May 13, 2008, 07:17:48 PM
We squeeze them in while the code compiles

Also, IIRC, for humans sleep doesn't burn that much less calories than being awake.  Mostly sleep seems to be for the brain to defrag, which takes a lot of calories.  The rest of the body apparently doesn't really care if it's just laying still and not moving or actually asleep.
Title: The State of the Simulation
Post by: goffrie on May 13, 2008, 09:31:36 PM
Hmm, you might know this, but http://xkcd.com/303/ (http://xkcd.com/303/)
Title: The State of the Simulation
Post by: Trafalgar on May 13, 2008, 09:42:09 PM
Does Spinner's strbody tie attack actually instakill? I thought strbody/fdbody was capped at 100 energy per cycle in either direction.

Title: The State of the Simulation
Post by: Numsgil on May 13, 2008, 09:55:40 PM
Quote from: goffrie
Hmm, you might know this, but http://xkcd.com/303/ (http://xkcd.com/303/)


 Yeah, XKCD is on my list of webcomics that I check out every morning.
Title: The State of the Simulation
Post by: EricL on May 13, 2008, 10:00:06 PM
Quote from: Trafalgar
I thought strbody/fdbody was capped at 100 energy per cycle in either direction.
It is.
Title: The State of the Simulation
Post by: The_Duck on May 14, 2008, 09:31:01 PM
Quote from: goffrie
Hmm, you might know this, but http://xkcd.com/303/ (http://xkcd.com/303/)

and http://xkcd.com/203/ (http://xkcd.com/203/)
Title: The State of the Simulation
Post by: Trafalgar on May 14, 2008, 10:46:52 PM
For a time, my sim consisted of a vast desert with one or two colonies of alga antiviral. Most things which teleported in never found the food. Now, however, it seems to have been taken over by shrinking violet 2.

I have to say, I really don't like them. I don't think veggies should be running around in a chain and killing everything they find with shots and tie attacks.  

Or, to have any chance of my normal bots competing, I'd have to mark them as veggies.

The worst part is that there are no normal edible veggies anywhere due to there already being 60+ killer mobile veggies in the sim.

(Does anyone know if these are actually the original form of these bots, or if their genomes were replaced by another bot? I haven't seen them in the IM before, they just showed up a couple hours ago here.)
Title: The State of the Simulation
Post by: EricL on May 14, 2008, 11:14:32 PM
Quote from: Trafalgar
(Does anyone know if these are actually the original form of these bots, or if their genomes were replaced by another bot? I haven't seen them in the IM before, they just showed up a couple hours ago here.)
I wrote SV and SV2 to demonstrate exactly this - that veggies are too powerful.  They can compete by simply runnign away and starving a sim.  They owned IM a month or two ago and I just re-introduced them.  For fun...    They're quite short and simple, they don't mutate, they use omin-eye vision and their defneses arn't bullet proof.  You should be able to get close enough to eat em.  Seasnake 1.3 seems to live off them okay.  But they sure don't act like veggies....

I've long made the argument that there should be some penality for being a veggie and getting all that free nrg from heaven such as no movement or vision or similar but have constantly been overruled by the evo veggie crowd.  There are several topics on this in the suggestiosn forum where I make detailed suggestiosn as to what to do.  Again, I have been consistantly overruled...  This is one reason both my IM sims have the sun turned off...

Oh, and if they are running around in chains usign ties, they are no longer origianl.  They have been infected with Seasnake DNA.
Title: The State of the Simulation
Post by: Trafalgar on May 15, 2008, 12:59:39 AM
Quote from: EricL
I've long made the argument that there should be some penality for being a veggie and getting all that free nrg from heaven such as no movement or vision or similar but have constantly been overruled by the evo veggie crowd.  There are several topics on this in the suggestiosn forum where I make detailed suggestiosn as to what to do.  Again, I have been consistantly overruled...  This is one reason both my IM sims have the sun turned off...

What if the veggie cap was per species?

And yeah, they must be infected with seasnake DNA, because they're anything but easy to kill.
Title: The State of the Simulation
Post by: Moonfisher on May 15, 2008, 07:36:44 AM
You could either increase costs severely for veggies to balance them out.
Or you could disable energy gain while evggies are moving or shooting and such, only give energy when they're doing nothing (Or virtualy nothing).
Or just require a veggy to fix itself before it can gain any energy, probably balance it so you need to be fixed for a while to get the full amount of energy per cycle.

Lots of options... but it's obvious that veggies are overpowered it's just a normal bot that gains 40 nrg per cycle instead of loosing around 1-2... it does have a pop cap but still...
Title: The State of the Simulation
Post by: Testlund on May 15, 2008, 02:24:27 PM
Quote from: Trafalgar
I have to say, I really don't like them. I don't think veggies should be running around in a chain and killing everything they find with shots and tie attacks.

That's why I only give my veggies 1 nrg/cycle and have day/night cycles set to 32000, and have vision turned off. The difference between an autotroph hunter and heterotroph hunter should be insignificant. The reason why veggies are the only ones that survives in my sim is because of costX that I need to use to prevent my sim to choke with bots. A hunter that can use it's energy a little more strategically, like shooting only when there's something to shoot at, and not constantly swims around aimlessly shouldn't have any problems surviving in my sim. On the contrary such a bot would most likely take over my sim completely.
Title: The State of the Simulation
Post by: Peksa on May 15, 2008, 04:20:29 PM
How would costs per species sound?
Title: The State of the Simulation
Post by: Testlund on May 15, 2008, 07:34:58 PM
How about an extra checkbox on the costs tab; include heterotrophs in costx? If just one of them is checked then increases/decreases in costX will only apply for the one checked.
Title: The State of the Simulation
Post by: shvarz on May 16, 2008, 04:46:14 PM
Quote
I wrote SV and SV2 to demonstrate exactly this - that veggies are too powerful.

Well, they are too powerful only under F1 conditions. Simple tweaks in costs and feeding take their advantages away. SV and SV2 were never able to get even slightly successful in my sim, because I feed veggies per kilobody point and have no limit on the number of veggies, only on total energy in the sim. SV2 just explodes into a hundred tiny bots that don't possess enough body to get energy for running and surviving the "night". So they die out. Simple as that.
Title: The State of the Simulation
Post by: Trafalgar on May 16, 2008, 06:10:33 PM
Wow, Spinner seems to have gone extinct.
Title: The State of the Simulation
Post by: Numsgil on May 20, 2008, 09:12:19 PM
I have a couple hundred sea snakes in my sim.  They all seem to be only two bots long.  Is that on purpose, or is it a mutation or something?
Title: The State of the Simulation
Post by: EricL on May 20, 2008, 09:59:35 PM
Quote from: Numsgil
I have a couple hundred sea snakes in my sim.  They all seem to be only two bots long.  Is that on purpose, or is it a mutation or something?
They have mutated in a round about fashion.  Seasnake itself is marked not to mutate, but it shoots it's single gene as a virus, which can mutate in other host bots and in turn, get shot back into seasnake hosts.  As long as the DNA length stays at 1727, original seasnakes won't kill mutated versions.  There hasn't been much competition lately, so there's nothing to preserve the MB.  The snake genome is drifting.   Slowly but surely, evolution finds a way....
Title: The State of the Simulation
Post by: gymsum on May 21, 2008, 12:38:37 AM
I think I've found your Seasnake's advesary... Run the PLankton with point, copy and insetion mutations only and I find that the seasnake often fails to keep its dna preserved.
Title: The State of the Simulation
Post by: shvarz on May 28, 2008, 08:52:21 PM
For the last two days I've had no luck connecting to the IM server   Not a singe time....
Title: The State of the Simulation
Post by: shvarz on May 30, 2008, 04:56:24 PM
Now it's working!

But anyway, nothing fun comes in   I get tons of SeaSnakes from Eric, but they can't compete with my Alga minimalis, even when ALL costs are turned off.  This really sucks, as I'm seeing a lot of diversity in behavior evolving even within my own sim. It goes through stages, I had "sit next to shape and fire" bots, then "just run and shoot" bots, then "spin and shoot" bots, and now they seem to move toward "run in circles" bots. Remember circumgyrans? Or whatever they were called...  Well, I have this stuff evolved in my sim. See the attached pic.
Title: The State of the Simulation
Post by: Numsgil on May 30, 2008, 05:04:27 PM
heh, that's neat.
Title: The State of the Simulation
Post by: shvarz on May 30, 2008, 05:44:23 PM
And just to illustrate the level of diversity that I see, here's a pic of two more sub-species (notice the difference in color) that exist in the same sim at the same time with the stuff I posted above.  If only we had more coherent sims connected to each other...
Title: The State of the Simulation
Post by: Numsgil on May 30, 2008, 05:52:17 PM
I have about a half dozen computers I'm going to connect as soon as IM can be left unattended for weeks on end without breaking.
Title: The State of the Simulation
Post by: asterixx on June 09, 2008, 09:30:25 PM
Well, after about 850 hours, my zerobot sim collapsed, and I tried bringing it back to life but they kept going extinct; I think I can blame that on there being no family recognition, although the last bot alive wasn't the "bully/mass murderer" as I had expected so I dont know. I'm going to abandon that project for awhile and just try tinkering with internet mode: Is it functional?
Title: The State of the Simulation
Post by: Numsgil on June 09, 2008, 11:13:24 PM
It is, but may need to restart the program from time to time if the connection goes bad.
Title: The State of the Simulation
Post by: shvarz on June 12, 2008, 02:06:49 PM
Backuping your bots works!
I have local teleporters that save and load bots to/from local folder.
I originally had a system of Duplo/Alga shellular that was evolving for a long time, several months. I had >1200 bots backed up during that time. Then Eric introduced SeaSnake. The snake itself could not make it in my sim, but Alga minimalis infected with Snake genome could. They came in and wiped out my previous sim completely. For the last month I only had these Alga minimalis in my sim.  I also changed conditions a lot by adding shapes, modifying energy flow and tweaking costs.
Just an hour ago the local teleporter brought in a Duplo that was saved over three months ago and that Duplo is doing so great, it multiplied to >600 bots.  This is fun!  Let's see what happens now.