Zerobot evolution prize for conditional logic

[EricL]

make sure those shepards are marked virus immune....

Err... That reminds of me of the other thing I wanted to mention that a virus immune bot can still shoot viruses via info shots. I've had several of the Shepards causing each other to fire viruses into the zerobots. Maybe we could add a check box to make the bot unable to fire viruses, in addition to the virus immune one we have already?

An excellent idea.  Consider it done for the next drop.

[Trafalgar]

Not to my knowledge.  The prize is certainly still available.

I posted this on aug 11: http://www.darwinbots.com/Forum/index.php?...p;#entry1371550

I had uploaded the bot's DNA in a zipped txt file here: http://shadowlord13.googlepages.com/bigzer...aimdx_3aims.zip (that's still a valid link)

The zerobot-descendants I wrote about in that thread had evolved pacing back and forth. If I analyzed the zerobot's DNA correctly,  it was setting .dn to -15 normally, but changing it to 365 whenever angle(0,0) < 461.

I described the approach I was using to evolving them in a post on Aug 4 and another on the 9th, in this thread (both of those were prior to the conditional evolution mentioned on the 11th): http://www.darwinbots.com/Forum/index.php?...amp;pid=1371468

I don't know if I posted the DNA for any version of my shepherd bots (I was modifying them as I went to try to encourage different directions of development) on the forums anywhere, but I still have the saved sims, and I can post the semi-unreadable DNA or readable-but-not-DB-code version of the last version of the shepherds.

This is a screenshot showing (part of) the list of saved sim files sorted by date (that's only showing the most recent 1/3 of the saved sim files): http://screencast.com/t/6XvlVQA6n

I can zip and upload any of these you would like, but I'd need to know which. I'm attempting to 7-zip all the ones from bzb4 and up now, and it's going slow, and who knows how big the 7z file will be.

(bigzerobots-1.sim, for reference, was on 7/13/2007)

I've skimmed through the changelogs, and if I understood them correctly, it sounds like some of the changes to cond stack stuff would break the conditional-behavior of the bots that I had evolved. (There are three conds which together determine whether the last gene runs, and it is that gene which gives rise to the conditional behavior)

Edit: The bzb4 sim files and all later ones were compressed by 7zip to 1.63 MB. I've uploaded the 7z file: http://shadowlord13.googlepages.com/main_bzb4to13_sims.7z

I think the bots in those sims evolved in 2.43 and 2.43a (I typo'd the version on the aug 11 post as "4.23a" and never noticed until just now).

[EricL]

I tend to take a month or two off in the Summer time and go sailing with my family, so I was AWOL last August and never saw your post.  It's late here now, but I'll take a detailed look tomorrow.  At first blush, I'm happy to award you the prize as well as keep it open for a second winner on modern builds, but let me read through it in detail tomorrow.

[Prsn828]

So I am 15 hours into my first zeroBot sim... and they evolved to reproduce on their own!!!  Right now they are veggies and I will be increasing the costs until I have the 5-10 best, but after that it will be time to throw them into a new sim as regular bots.  The sim started out with 5 50 zero bots and 5 100 zero bots, but eventually the 100 zero bots wiped out the 50 zero bots.  I wonder if this is how it usually happens... anyway, I'm happy to say that I am pleased with the progress they made in the last 15 hours.

[Numsgil]

Congratulations   Keep up posted.

[ikke]

The sim started out with 5 50 zero bots and 5 100 zero bots, but eventually the 100 zero bots wiped out the 50 zero bots.  I wonder if this is how it usually happens...

It's about combinations and permutations. The 100 BP variant had more possibilities of mutating BP into something useful. So yes I do think this will happen usually

[Numsgil]

I'd like to see a zerobot sim sometime with bots with thousands of bps of 0s.  Would be interesting to see if more interesting things develop that way.

[ikke]

I'd like to see a zerobot sim sometime with bots with thousands of bps of 0s.  Would be interesting to see if more interesting things develop that way.

Under default conditions the only thing happening is deletions decreasing genome length. The bot sin my sims started out with 1000 BP and shrunk without uniquely exiting things happening.

[Numsgil]

They'd at least have to start reproducing before genome length decreased.  Unless maybe the bots found .delgene or something.

[ikke]

They'd at least have to start reproducing before genome length decreased.

Yep. A large genome does not automagically translate into complex behaviour. A complex environment is needed for that. Eating minimalis does not qualify.

[intron]

Are initional virus infection or introns* allowed?

* like
cond
start
*.thisbp *.thisbp 8 add 0 .dnalen rnd insert
stop

[Numsgil]

I don't think that would qualify as a zerobot.  But it would probably give better results, so don't be afraid to give it a try (btw there is no *.thisbp.  You'll have to make do with .thisgene).

[Houshalter]

Is there an insert command? I didn't think there was.

[Numsgil]

No, you have to use a virus.

Actually, I should have linked this page on making viruses.

[Houshalter]

I made this virus a while ago to spread genes throughout a population like Horizontal gene transfer. It kind of works, just make sure the veggies are immune or they mutate it to be destructive, then spread it like wildfire.

[div class=\'codetop\']CODE[div class=\'codemain\' style=\'height:200px;white-space:pre;overflow:auto\']start
*.vtimer 0 =
3 rnd 1 =
*.thisgene .mkvirus store
not
*.genes rnd .mkvirus store
*.vtimer 1 =
*.eye5 0 > and
*.nrg 15 div .vshoot store
*870 0 =
*.genes 870 store
*.genes *870 >
*.genes rnd .delgene
*.robage 0 =
100 rnd 871 store
*871 66 >=
870 dec
*.robage 1 =
*871 33 <=
870 inc
*.robage 1 =
0 871 store
*872 *.thisgene !=
*872 .delgene store
*.thisgene 872 store
stop