Author Topic: 2.4.8  (Read 2784 times)

Offline shvarz

  • Bot God
  • *****
  • Posts: 1341
    • View Profile
2.4.8
« on: November 07, 2005, 02:37:55 AM »
IMO

I thought we agreed on that.  Nums, why are you still working on that?  Is there anyone here who is interested in that?  I thought we had people who wanted a stable 2.37 and we had people who wanted a brand-new version.  Did I miss something?
"Never underestimate the power of stupid things in big numbers" - Serious Sam

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7738
    • View Profile
2.4.8
« Reply #1 on: November 07, 2005, 11:48:40 AM »
The C++ version is 100% just copied and redone from the VB source.  Any time I spend on the VB engine code isn't necessarily wasted.  GUI is a different matter...

I need people to use 2.4 so I have a stable version to compare against when the C++ version is done, so bugs are highlighted.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7738
    • View Profile
2.4.8
« Reply #2 on: November 07, 2005, 01:08:00 PM »
Quote
I haven't noticed the color change for mutations in 2.4.8 ...
but then only did a quick test run.
It's very subtle change.  You'll need at least 3 mutations to see any difference at all, and even then it will be only a slightly different hue.

Quote
could it be mutations are not enabled by default?
will play more as I find time and see how she goes.

Possibly.  I thought I changed what they defaulted to but it wouldn't be the first time the program goes left when I distinctly remember telling it to go right.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7738
    • View Profile
2.4.8
« Reply #3 on: November 07, 2005, 01:09:16 PM »
Quote
re mutations and color changes:

ran a long sim and could finally detect some slight color changes ...
however, there were many mutations and I would expect to see many more color shifts.
picking one and taking a look at robt data I found:
although only 1 Overall mutation was indicated ...
there were actually 6 mutations that had occured:
the dna details:

insertion changed the: to the logic operator: and at postion 56 during cycle 8276 ...
which then goes on to describe 5 others.

the same with other bots I checked with mutations ....
only a few of the mutations show up in the main robot data chart ...
another listed 4 overall mutations ... but there were actually 7 ...
well ... 7 were indicated in the detailed dna data anyway ...
which of these is correct would only be a guess.
I didn't go into the dna to see if all the changes listed had been made.
Sometimes the mutation details will list two events because the event was two long I think.

See if any of the mutations were occuring on the same cycle.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7738
    • View Profile
2.4.8
« Reply #4 on: November 07, 2005, 02:51:18 PM »
I haven't even touched the code for graphs, so you're on your own there...

I don't think you're understanding how use NewMove works.  Older bots with use NewMove in their DNA will function oddly.  Remember early versions of 2.4 where baby animal minimalis's would vibrate back and forth?  That's what they'd be doing.

I'll check out your settings...

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7738
    • View Profile
2.4.8
« Reply #5 on: November 07, 2005, 03:53:19 PM »
Quote
talked about how it wasn't fair ... with the new physics ...
that the old bots movements were multiplied ...
and therefore to add 'use NewMove' so they would not be.
that IS what you said ... or at least that is how I interpreted it.
That's right.  But older bots won't function properly unless their movements are multiplied.  So adding use NewMove to an older bot will make it function oddly.

How was my explanation half assed?  I think I explained it to a T.  bots without use NewMove will have thier movements multiplied by their mass as a handicap.

Quote
when should it be used and when not?

Newer bots are encouraged not to use it, and instead to manually modify accelerations as their mass increases.  ie: Have use NewMove at the top of their DNA.

It's designed so that older bots are plug-n-play.

Quote
please be as specific and as detailed as you can ...
don't just throw this shit out there like everyone is supposed to
be able to read your mind.
I'm telling you man ...
you aren't worth a shit at explaining things ...

People who actually modify DNA seem to know what I'm talking about.

I don't get whats so hard about this really.  It's not some mystic new command.  It auto multiplies mass by the .up commands if it's missing, so older bots will work right.  It should be a mind bender.

Let me explain it in terms more appropriate for your mental faculties:

no use NewMove mean {.up = .up * mass}
« Last Edit: November 07, 2005, 03:54:29 PM by Numsgil »

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7738
    • View Profile
2.4.8
« Reply #6 on: November 07, 2005, 03:58:59 PM »
Quote
ran the Swarm setting in 2.4.7 ...
no problem ...
the veggies work ok and don't go nuts trying to reproduce as they do in 2.4.8
also ... in 2.4.8 ... the bots, T_Preservans ... seem to go right thru the veggies ...
whereas in 2.4.7 ... they hit and bounce off of them ... and interact as they should.
Haven't touched collisions at all.  Veggies go nuts trying to reproduce is probably a mutation.  Try running without mutations.

Again, allow me to condescend:

button in mutations panel:
'mutations enabled' means mutations are enabled.

'mutations disabled' means mutations are disabled.

I'll check out T_Preservans.  I don't imagine it's really cut out for modern DB anymore...  It's been obsolete for quite a while.

Try DIN, Dominator Invincibalis, Icarus, et al.

Offline shvarz

  • Bot God
  • *****
  • Posts: 1341
    • View Profile
2.4.8
« Reply #7 on: November 07, 2005, 05:42:47 PM »
Quote
I need people to use 2.4 so I have a stable version to compare against when the C++ version is done, so bugs are highlighted.

OK, maybe I understand that.  But if 2.4 is buggy itself, then is that going to be helpful?  And just imagine the confusion any newcomer will experience due to a release that seems to be just a minor upgrade from 2.37, but in fact is a whole different program that is created only for your convience.  Are they supposed to d/l that?  And will you personally teach them to use "no newmove" or whatever fix you used to make older bots work?

I am not trying to bash you or anyone else on the board, but really we need to step back and take a look at the project as a whole.
"Never underestimate the power of stupid things in big numbers" - Serious Sam

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7738
    • View Profile
2.4.8
« Reply #8 on: November 07, 2005, 07:25:18 PM »
Okay, let me be very explicit...

use NewMove is built such that older bots will work because they don't have it.

hence the 'new' in new move.  If it was meant to be something older bots used, I'd have called it OldMove.

So there's nothing to teach new users.  They can happily chug away with older bots and everything'll be just fine.

You know what, forget I said anything about use NewMove.  I don't mean that as a hopeless 'I give up' sort of thing, I mean it sincerely.  Pretend I never mentioned it.  It's for new bots anyway, and since there aren't any new bots, it's all a moot point.

2.4 is not really that buggy.  I'd say it's on par or better than 2.37.4.  It's really not a 'work in progress', it's becoming a stable, stand alone version all its own.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7738
    • View Profile
2.4.8
« Reply #9 on: November 08, 2005, 10:32:46 AM »
I just fixed a stack bug, set up Darwinbots to run with DIN in evo mode all night on default settings (slightly tweaked from 2.4.8).  I came back this morning and they've all shrunk (because shots return fixed amounts of energy instead of based on body) and their population exceeds 1000.

I'll upload 2.4.9 with those minor changes later today.

Your problem Griz is that I could fix 100 out of 101 bugs, and I'll get no gratitude for my efforts.  That's very discouraging on both sides.

Find me where I said use NewMove did exactly the opposite of what I just said.  I dare you.  That might be how you read it, but I can't be held accountable for your lack of reading comprehension skills, now can I.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7738
    • View Profile
2.4.8
« Reply #10 on: November 09, 2005, 11:45:54 AM »
Quote
yeah ... ok.
like I said ...
you're never wrong, never mistaken. ;)

I'd agree except I know you're being sarcastic.