Messages

46

Bugs and fixes / 2.36 bugs

« on: April 02, 2005, 12:47:48 PM »

Just tried 2.36. I lifted a working sim from 2.35 and ran it in 2.36.  Major differences noted.

1. Eye values seem to be constrained. Bots do not "see" as far ahead as in 2.35 nor as wide. Major differences noted in bot's ability to see then maneuver to a distant veggie.

2. Eye values change more dramatically.  In testing 2 bots approaching a veggie the first eye values in eye5 were 28 and 22 respectively. No turning was involved. This was a straight shot test.  I also noted in each case that eye5 value jumped dramatically from 33 and 38 to 55 and 60 in one cycle. 2.35 seems to be different in the inverse distance formula used for eye values.

3. At eye5 = 77 bot is still more than 1 cell length from the veggie. At value 55 the bot is about 2 cell lengths from target.

I had a major problem with veggies getting killed. The bots are trained to not kill veggies. I went into the Sim Parameters & Options menus and found the following:

4. In Costs screen, Energy Exchange - Fixed and Proportional both at 200. The sim I lifted from 2.35 did not have these values. Value arrows (up and down) on either option only changes value in Proportional window. Once set using keyboard and restart sim, values reverted to 200.

5. In Recording - Autosave - Value arrows on Whole Sim change Best Bot time.
                                          Value arrows on Best Bot change Whole Sim time.

6. Veggies receiving no energy. A_Minimalis losing about 1 point every 2-3 cycles. A_Minimalis_flotus loses 1 point about every 100 cycles.

7. With all eye values at 79, bots shots trail off to the left missing the target. Eye5 orientation appears to be off.

I cannot speak to the more advanced functions yet. Seems the most basic functions are significantly different between the two versions.

Off the wall juxtaposition: As a person of average heigth stand up and imagine a circle with a diameter equal to your heigth. Imagine this is Earth. Now take one finger and inscribe the circumference of Earth's surface. With the flat underside of your finger inscribing the surface the topside, one finger width from the surface, is the altitude the ISS and the shuttle fly. "Outter space" it ain't. But it is the first step.

-P

47

DNA - General / Racial memory

« on: April 01, 2005, 03:10:31 AM »

Thanks, Bots.

Neither Alpha nor Delta are intended nor destained to be plants.  Right now I'm looking at populating the sex-trait values through the birth tie.  That should suffice until (hint) .sexrepro is working (end hint).

From Sir Nums:

That's why I wanted to add it. Junk DNA allows mutations to accumulate without effecting the phenotype immediatly. Presumably the cost/benefit ratio of pruning DNA of Junk DNA is just too high, so cells don't really try.

Or there is utility to that 80+ percent of our genome that we label as junk and the cells [you]must[/you] keep it around. I'm sure the designated smart guys will figure this out in not too long a time.

If you add some "junk DNA" space to DB can I surrepticiously abscound with some of it for real DB DNA?  Please?

Reminds me of David Sarnoff when talking to RCA's lab people about building a computer to compete with IBM.  

Four 4-bit processors in parallel with a staggering 512k of that new silicon memory stuff and a bank of new 30" iron/oxide platter drives for 5 meg of data each.

"No one is ever going to use that much space and power!"   Right.

Gimme the space and I'll fill it.

Trivia tidbit: The first small (8") individual hard drives were called "winchester drives" in the industry since IBM's marketing group designated them the WN-3030 drive. Correlated with the old Winchester 3030 rifle. We play in an industry built by cowboys.

Cowboys are special
With their own brand of misery
From bein' alone too long.
You could die in the cold
In the arms of a nightmare
Knowin' well that your best days are gone.
Pickin' up hookers
Instead of my pen
I let the words of my youth fade away.
Old worn out saddles
And old worn out memories
With no one and no place to stay.

-Willie Nelson - My Heroes Have Always Been Cowboys

48

DNA - General / Racial memory

« on: March 31, 2005, 04:40:14 PM »

Thank you both.

FYI.  I only checked two locations (980, 981) but I did console check 4 newborns one cycle at a time from 0 thru 89. Mom's memory values did not transfer.

I've gotten around it, for now, letting the initial rnd function determine the value for the male tail length and female tail preference. I suppose I could communicate mom's values across the tie. I need to look at that.

Thanks, again.

Just off the wall: I like the name "junk DNA." This speaks volumes about us crazy H. Sapiens. We can't see the utility, if there is any, so we assume it has none and is thus junk. In humans there are lots of old bacteria and viral remnents in there, btw. I like, better, McFadden's view that it's leftover stuff from eons of development no longer used and may be "sleeping" awaiting the right chemical alarm clock. Gould speculated this could be the expansion ground of new features for some future H. Intellectulus. Some junk.

Enough.

Bye, y'all.

-P

49

DNA - General / Racial memory

« on: March 31, 2005, 11:53:32 AM »

Back again, folks.

I tested locations 980 and 981. These do not transfer. I will try other locations as I have time.

FYI. Birth tie separates at age 89.

I am using DB2.35. Is racial memory active in this version?

I'm almost done with Alpha. Sex-based (simulated, since .sexrepro doesn't work yet - hint) with separate male/female behaviors, female mate selection, and sex trait inheritance (in the absense of .sexrepro gene shuffle - hint) through racial memory (if I can get that working) all contained in one DNA set (right up against the DNA length limit). Finally got the eyes working well enough (no math in conditionals - man wasn't that a headache - but it does so state in the documentation, damnit) and movements smoothed out.  Alpha is a pretty little bird that eats and makes babies. No protections (except dodging bad guys) and is intended for Evo Sims not for combat.  I've had to combine some gene processing so as to stay within the length limit but there are still plenty of genes for a good sex shuffle (if we had one - hint).

When can I have a real working .sexrepro?

Thanks, y'all.

-P

50

DNA - General / Racial memory

« on: March 30, 2005, 08:25:38 PM »

Since the "search" function is down I'll do it this way.

In some post in some topic on one of these boards I remember reading about certain mem values (in the 900s) transferred to baby after the 20th cycle of the birth tie. I can't find it.

Would some kind person please repeat the info? What memloc's?

I promise I will, this time, copy and paste the info into my own documentation for future reference.

Thanks, as always, people,

-P

51

DNA - General / Math stack issue

« on: March 29, 2005, 02:04:18 PM »

Sorry to inform you of this but.......

MATH FUNCTIONS DON'T WORK IN CONDITIONS!

 Say what? :shoot:

 :burnup:  :banghead:  :cuss:  :cuss:  :tantrum:  :shoot:  :bash:  :cuss:  :rant:  :pokey:  :bash:  :plzdie:  :shoot:

52

DNA - General / Math stack issue

« on: March 29, 2005, 01:03:43 PM »

I feel so [you](*&% [/you]incompetent.

How could something so [you]&*^%[/you] simple cause me so much [you]&^%&#@[/you] headache!

Eye values:  0 0 0 0 0 0 24 24 24

Gene:

'
'----- #21  Something to the right - turn
'
cond
  *.eye5 *.eye4 add *.eye6 add 0 =
  *.eye7 *.eye9 add *.eye1 sub *.eye3 sub 0 >
start
  100 .aimdx store
stop

Unless I have a totally bogus view of how the [you]&(*^^#[/you] math stack works both of these [you]&^$@@%*)[/you] conditionals should return me a [you]*&%$@[/you] "true."

So why the ever lovin [you]&^%$[/you] won't this stupid [you]%*&^[/you] gene fire?

No other genes executed during the cycle (but this one was supposed to, the [you]&^%$[/you] thing). So what the [you]*&^$@#*((&%[/you] is going on here?

Embarassingly yours,

-P

53

DNA - General / Console question

« on: March 29, 2005, 12:19:43 PM »

Well, let's see.

• 12
• 23
• 34
• 56
• 78

Seems to work fine. Am I not the only one with headspace and timing problems?

Shouldn't the [list=a] at the start just be ["list"]?

(Had to put the quotes , or something, in the brackets to keep list from activating.)

Maybe we should try this (write the list then bookend it with the list commands):

1. I like chocolate pudding.
2. I like the outter edge of the universe.
3. Therefore the outter edge of the universe is made of chocolate pudding.[/li]

Now this works right fine. I think maybe it's the list=a thing at the front, PY.

54

DNA - General / Console question

« on: March 28, 2005, 05:28:16 PM »

The DNA is parsed at the start of the cycle. This places all of the values into the robot's memory cells.
The console displays them at this point so when the console tells you that eye5 sees 64, that is exactly what the robot sees right now.
The memory values haven't been processed and converted into actions yet.

Now, you see, I read this as just the opposite of what I asked. Headspace and timing on the newbie's part. Let me get verbose so I understand:

Robot DNA manipulates memory block  >>---------------------- cycle start

Robot memory block passed to program for execution            

Program executes then returns memory block to robot   <<---- cycle end
<<------------------------------------------------------------------console view
Robot DNA manipulates memory block

Robot memory block passed to program for execution

Program executes then returns memory block to robot
<<----------------------------------------------------------------- console view
Robot DNA manipulates memory block

Is this correct?

Don't you just love ASCII as a drawing tool?

-P

55

DNA - General / Console question

« on: March 28, 2005, 02:04:48 PM »

Howdy, y'all,

Question on what I'm look at on the console.

If my understanding is correct the memory block I can see in the console is the one built [you]after[/you] program execution [you]before[/you] the DNA acts on it. Yes?

Much obliged,

-P

56

Off Topic / Happy Birthday Purple Youko

« on: March 28, 2005, 01:45:15 PM »

Happy happy, PY, you youngin.

Just so's you know, it don't get no easier from here on.  Enjoy while you can.

-P

57

DNA - General / Exploding Babies - Help!

« on: March 25, 2005, 05:38:46 PM »

Shvarz!  You sweetheart! My hero!

I didn't even look at that stupid gene since, I donno, didn't seem important in the reproductive process.

Thank you, thank you.

No need for a DNA compiler (though it would be nice). Just some additional headspace and timing on the bot maker's part.

All is right with the universe.

-P

58

DNA - General / Exploding Babies - Help!

« on: March 25, 2005, 02:27:54 PM »

I am lost. My head aches. I cannot see the issue through my tears.

My Delta is down. All of her babies explode on cycle 1.

I've read (and experienced) that this is usually a coding problem in the DNA (missing start, stop, etc.). I can't find it.

Maybe momma is shooting the baby? I've no-op'ed, coded around, conditionalized mom's feeding genes to no avail.

I no-op'ed the "birth tie" genes, movement genes, everything I could think of.  Sadly, Delta's babies still poof like a bullet in a campfire.

She was working and I've compared old version with this one and can't see where I've  #$*^'ed up.

To run a properly controlled sim I need my Alpha-1 (will not mutate), Alpha-2 (allowed to mutate) and selection pressures (food & predator) all well adapted to their present starting niche. I did not like Delta's hunt routine and made copious modifications. Now she's done got broke and I can't see the fix.

Anyone with some time and inclination, please help.  She is attached.

Be gentle with her.

Heartbrokenly yours,

-P

59

DNA - General / 5 Questions re DB opns

« on: March 25, 2005, 02:02:03 PM »

Thanks folks, again.

I was just wondering how this feature worked in 2.35

It is not a major issue, PY, and no need to spend time searching through old code.

The issue is interesting, though.

The "goal" of evolution, if it can be said to have one, is procreation. Dawkin's Selfish Gene, if you will. Number of offspring is the prime key, of course. For programming efforts, though, the secondary and tertiary sort selection is a quandry.

Is longevity a proper second?  Is a slow breeder that lives long "better" than a fast breeder with limited life? I donno.

Most cumulative mutations? Maybe. That is of interest. That is the reason for the simulation isn't it. Most mutations from the start of the sim and still tied for top breeder? That may indeed be "best."

My input: # offspring
              # cumulative mutations
              # cycles (age)
              any bot with "42" in the DNA

Well, maybe not that last one.

-P

60

DNA - General / 5 Questions re DB opns

« on: March 25, 2005, 02:30:51 AM »

Nums, PY, shvars,

Excellent, Gents. Excellent indeed.

One outstanding:  Best bot -- when multiple bots have the same # kids what criteria does the program use to select which one for autosave? The red one? Something with "42" in it since 42 is the answer to...well, you know.

As for the remainder:

Got it all. Everything is copacetic. I have seen the error of my ways. I am armed with copious correct information. I go forth now knowing that the true answers are mine. I am content.

For the moment.

Thanks, all, again.

-P