Darwinbots Forum
Bots and Simulations => Bot Tavern => Topic started by: maheshjr2000 on May 21, 2006, 01:28:06 PM
-
So far all ive seen in the autotroph section is algae(feel free to correct me if im wrong). So I wanted to create a plant autotroph. It would involve using ties to feed a central cell which would be responsible for most of the reproduction till another cell was pushed out of the "plant" and started reproducing itself. If you guys have any advice itll be most helpfull.
-
There have been a few multi-cellular plants designed but nothing with any real structure.
There has been talk of making plants with roots, trunks, branches and leaves but so far nobody has made one. It is certainly an area that could use some development.
-
Yeah, that sure whould be cool if you could develop plants that grows from the bottom and then make multicellular organisms that feed from the plants. ...and predators. If there whould be any advantage for a plant to grow in this environment it might evolve on it's own from single cellular algea maybe. ...if I keep running this evosim for a few years. *Sigh* Not very likely.
-
There's Plant v1.0 (http://www.darwinbots.com/Forum/index.php?showtopic=1019) in the Beastiary
And the associated Bot Tavern post (http://www.darwinbots.com/Forum/index.php?showtopic=1012) - There's quite a lot of ideas in there actually
Maybe once cell specialisation and metabolism are introduced we might se some plants with specialised parts, but for now all parts of an autotroph can photosynthesize
The idea of a multibot-veg is quite intreguing though ...
-
Well that bot only consisted of three cells. This one will be a lot more cells with each cell passing energy to the central cell responsible for most of the repro considering the amount of energy it gets. I tried to do that with ties but the result was alga_parasite. So this time I was thinking shots or an intelligent system of ties.
-
I know I may be double posting but I just got a new idea is there anyway to destroy a childs dna before it activates? Like make it an empty shell?
-
Not the very moment it's created. You could have the child delgene all its genes.
-
how long would that take?
-
A cycle for every gene, but the child would have to do it itself, so you'd need to add that condition into the genome of your species.
-
could it delete the delete gene?
-
Yes
-
how would I get it so that only the sons delete themselves and not the parent? because then I might have an idea for a multicell veggie.
it uses four sons to gather energy cus they are still technically veggies and once it has a certain amount of energy it pumps dna into the first son breaks the tie and sends him off to start a new plant.
-
First you need to specify some way to identify soon to be DNAless cells. Say, *50 = 1.
Then a gene like this (at the end of the genome) would delete all the genes in the bot:
cond
*50 1 =
start
.delgene inc
stop
-
so for the father cell I would make it so that location 50 was equal to 0 and for the sons it would be equal to 0
-
I assume that's a typo and you really meant:
for the sons it would be equal to 1
Which would be right.
-
Then how about something like:
cond
*.in1 77 =
start
1 50 store
stop
And so when a parent wants to give birth to a 'blank slate' baby it just has to put 77 in .out1 before reproducing
-
Numsigil youre right about the typo.
Elite thats a good idea! Originally I had thought to use the existing blank slates and insert dna in but that would make them old. However instead of the .in method I was going to use 2 different repro genes.
Also any ideas on how to accelerate the baby at birth?
-
One final question do the memory locations reset each cycle?
-
Depends on the location. So do, some don't, some are read only, some are write only.
-
Thanks guys! I should have the first version out soon .
edit: do birth ties count as part of numties?
-
as far as I remmember no
-
edit: do birth ties count as part of numties?
1) They do in 2.4X.
2) This is trivial to check via the bot console...
-
Also any ideas on how to accelerate the baby at birth?
This gene
32000 .tielen1 store '( 'Endy's gene from Elite's post somewhere )
I don't think it's going to work in 2.4+ so the work around would use the same gene but have to use (1 .tie) or (.tie inc) to tie to parent first.
When it comes to birth ties, overlaying and extending seems to be as competitive as overlaying and cutting anyway.
-
As we have discussed previously, this gene should not work in 2.37.6 either. If it does then it is a big bug that needs to be squished pronto.
Tielen should only work on hardened ties.
-
Don't think it's that much of an issue and it is nice for the bots to have a sort of "natural" method of deleting ties/avoiding attacks.
Off topic:
Weird how we're all on at the same time
Really strange since I'm sitting in Japan right now
Going to be gone for a bit again, I'll post when I return
Endy
-
awsome!!!!!
can you guys post here when its in the bestiary?
-
It should get posted in the starting gate first, plus you should see it under 'view new posts' if it ever appears.
-
There have been a few multi-cellular plants designed but nothing with any real structure.
There has been talk of making plants with roots, trunks, branches and leaves but so far nobody has made one. It is certainly an area that could use some development.
you should you made working ants so plants should be easy