Author Topic: Version 2.42.6 Released  (Read 8290 times)

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7725
    • View Profile
Version 2.42.6 Released
« Reply #15 on: June 15, 2006, 04:31:31 PM »
In 2.4 I made bouyancy a function of volume, as you would expect in real life.  The problem is the only thing bots have that effect their density, and hence bouyancy, is shell and body.
 
 If there isn't any fluid density, there won't be any fluid bouyancy.

Offline PurpleYouko

  • Bot God
  • *****
  • Posts: 2556
    • View Profile
Version 2.42.6 Released
« Reply #16 on: June 15, 2006, 04:54:48 PM »
Quote
In 2.4 I made bouyancy a function of volume, as you would expect in real life.
I wouldn't expect that at all.

Whales are huge and have neutral bouyancy at the surface and somewhat negative bouyancy as they dive to great depths as the air in their lungs gets compressed.

Tiny little crabs sink like a rock.

Fishes of all sizes are able to float up and down at will by the use of swim bladders. That is what bouyancy is supposed to represent

Quote
The problem is the only thing bots have that effect their density, and hence bouyancy, is shell and body.
Not true. I gave them swim bladders in the form of the .setboy sysvar.

Quote
If there isn't any fluid density, there won't be any fluid bouyancy.
Fluid density in pond mode is assumed to be that of water. It is a fixed and completely irrelevent value and is constant with depth. We don't need it to be defined beyond that.
The default position of all bots of any size is neutral bouyancy but this can be changed by internal adjustments (swim bladder) or external adjustments like the addition of shell which would obviously make them more dense.

Where exactly is the problem. This was completely working in 2.37.6 and what's more it was very simple and elegent.  
There are 10 kinds of people in the world
Those who understand binary.
and those who don't

:D PY :D

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7725
    • View Profile
Version 2.42.6 Released
« Reply #17 on: June 16, 2006, 03:17:52 PM »
Quote from: PurpleYouko
I wouldn't expect that at all.

Whales are huge and have neutral bouyancy at the surface and somewhat negative bouyancy as they dive to great depths as the air in their lungs gets compressed.

Tiny little crabs sink like a rock.

Fishes of all sizes are able to float up and down at will by the use of swim bladders. That is what bouyancy is supposed to represent

 All those things are on Earth and hence are also being acted upon by gravity.  The net movement in a liquid is a function of bouyancy and weight of the object (although bouyancy wouldn't exist in a zero G environment either).
 
 Hence, the net movement of an object is actually dependant upon density.  Bouyancy, however, is strictly dependant on volume.  (Force of Bouyancy = Volume * density of liquid * gravity, ie: Archimedean Principle)

Quote
Not true. I gave them swim bladders in the form of the .setboy sysvar.

 Which I couldn't figure exactly what to do with when I remade bouyancy.  I mentioned this at one point when I talked about adding a new constructable unit which would be massless but take up volume, to help counter shell which can be used as a balast (no volume, but adds mass), but I couldn't reach a concensus, so I put it on hold and apparently forgot about it.

Quote
Fluid density in pond mode is assumed to be that of water. It is a fixed and completely irrelevent value and is constant with depth. We don't need it to be defined beyond that.

 1.  Liquid can't be compressed, so it's density is constant at all depths.  Pressure, however, is not, which I assume is what you're confusing it with.  Since bots are assumed to be implosion-proof, pressure obviously doesn't matter.
 
 2.  It should be fairly obvious that you can define the density of the liquid since the advanced physics tab has a field where you can specify the density of the liquid.  In fact, if you read the toolbar in the advanced physics tab it tells you that density effects the force of bouyancy.  It also tells you the density of an average bot for comparison purposes.
 
 3.  "water" has no correlation with the bot universe because we have nothing to compare any of the metrics between the two environments with.  Density in real life is (kilo)grams / meter^3.  Density in DB is mass / twip^3.  It's comparing apples and oranges.
 
 
Quote
The default position of all bots of any size is neutral bouyancy but this can be changed by internal adjustments (swim bladder) or external adjustments like the addition of shell which would obviously make them more dense.

 Assuming a density of 1E-6 (which advanced physics tab suggests), this is true (minus the swim bladder, which I explained above).  1E-6 should be the density of body.  Adding shell would make you sink.  There's nothing you can add to decrease density, which is the problem I explained above.

Quote
Where exactly is the problem. This was completely working in 2.37.6 and what's more it was very simple and elegent.  

 It was perhaps simple but it did not correlate well with the rest of the physics.  By redoing all the physics, I tied all forces in together to create a central and stable way to integrate them.  By tying bouyancy in to fluid density, I correlated drag forces, bouyancy, etc. into a few parameters which correspond very well with real life.

Offline PurpleYouko

  • Bot God
  • *****
  • Posts: 2556
    • View Profile
Version 2.42.6 Released
« Reply #18 on: June 17, 2006, 06:53:25 PM »
Quote
All those things are on Earth and hence are also being acted upon by gravity. The net movement in a liquid is a function of bouyancy and weight of the object (although bouyancy wouldn't exist in a zero G environment either).
And so is "pond" mode. It represents a pond after all. You cann't have a pond unless you have gravity so the concept of a pond automatically assumes its presence.

Quote
Hence, the net movement of an object is actually dependant upon density. Bouyancy, however, is strictly dependant on volume. (Force of Bouyancy = Volume * density of liquid * gravity, ie: Archimedean Principle)
Except that archimedies principle is irrelevent when it is applied to aquatic creatures which swim in the water column. Their density is identical to the medium through which they swim as a default. They can change it with the addition of more dense stuff like shells or less dense stuff like swim bladders..

Quote
Which I couldn't figure exactly what to do with when I remade bouyancy. I mentioned this at one point when I talked about adding a new constructable unit which would be massless but take up volume, to help counter shell which can be used as a balast (no volume, but adds mass), but I couldn't reach a concensus, so I put it on hold and apparently forgot about it.
I don't even remember this but then I really didn't have a great deal to do with 2.4 most of the time

Quote
1. Liquid can't be compressed, so it's density is constant at all depths. Pressure, however, is not, which I assume is what you're confusing it with. Since bots are assumed to be implosion-proof, pressure obviously doesn't matter.
I'm not confusing anything. I stated that density of water is constant with depth and it is. I didn't mention pressure since any creature living at any depth will take on the density of its surrounding water. It is irrelevent as I said.

Quote
In fact, if you read the toolbar in the advanced physics tab it tells you that density effects the force of bouyancy. It also tells you the density of an average bot for comparison purposes.
That's all very fine and dandy except that in "Pond Mode" all that stuff should be totally over-ridden such that bot density == water density. Otherwise the concept of "Pond Mode" is a waste of space. Gravity is also a meaningless concept in "Pond Mode". All you need to know is that it exists. Any change in its magnitude will only serve to change the pressure acting throughout the whole system and since the robots and the water exist in a complete balance, nothing will change at all.

Quote
3. "water" has no correlation with the bot universe because we have nothing to compare any of the metrics between the two environments with. Density in real life is (kilo)grams / meter^3. Density in DB is mass / twip^3. It's comparing apples and oranges.
The exact density is utterly irrelevent. The math to calculate it is utterly irrelevent. The differences between the DB world and the real world are also utterly irrelevent.
There is only one thing here that is an issue of any kind.
"Pond Mode" was created to simulate a POND or TANK with all assumptions relating to water and bot density automatically implied. Equally, in "Pond Mode" the Y axis of the screen can only be equated to up and down in the vertical plane so many other assumptions are implied also.

Quote
Assuming a density of 1E-6 (which advanced physics tab suggests), this is true (minus the swim bladder, which I explained above). 1E-6 should be the density of body. Adding shell would make you sink.
Yup. That's right.
Quote
There's nothing you can add to decrease density, which is the problem I explained above.
Except that there quite obviously IS.
Fish do it all the time.
It's called a SWIM BLADDER. Fish are able to manipulate their density to float up or sink down at will. Plankton do it too. As do krill and many other aquatic creatures.

Quote
It was perhaps simple but it did not correlate well with the rest of the physics. By redoing all the physics, I tied all forces in together to create a central and stable way to integrate them. By tying bouyancy in to fluid density, I correlated drag forces, bouyancy, etc. into a few parameters which correspond very well with real life.
It didn't need to correlate with much of the other physics. It was designed to override them.
Frankly I have never really liked the physics of 2.4 much. That's another reason why I have still not made the switch.

The fact is that in almost every way that matters to me, 3.37.6 is still superior. 2.4 is getting better but it just lacks all the things that make DB fun for me.
There are 10 kinds of people in the world
Those who understand binary.
and those who don't

:D PY :D

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7725
    • View Profile
Version 2.42.6 Released
« Reply #19 on: June 17, 2006, 10:31:31 PM »
Quote from: PurpleYouko
Except that archimedies principle is irrelevent when it is applied to aquatic creatures which swim in the water column. Their density is identical to the medium through which they swim as a default. They can change it with the addition of more dense stuff like shells or less dense stuff like swim bladders..

 This is not exactly true.  Most fish are not in equilibrium density with their environment.
 
 Swim bladders are not found in all fish.   Many cartilaginous fish, such as sharks, can control their depth only by swimming; others store fats or oils for this purpose.
 
 
Quote
I'm not confusing anything. I stated that density of water is constant with depth and it is. I didn't mention pressure since any creature living at any depth will take on the density of its surrounding water. It is irrelevent as I said.

 I'm sorry, I read consistant instead of constant.  Changed the meaning of the whole sentence.

Quote
That's all very fine and dandy except that in "Pond Mode" all that stuff should be totally over-ridden such that bot density == water density.  Otherwise the concept of "Pond Mode" is a waste of space.

 At present pond mode implies only the way in which light is distributed in the sim.  What if I want to run a sim with a gradient for the vegs in a total vacuum?  I'm certain I mentioned this when I split corpse mode and pond mode apart into their most core features.  I removed the hooks to the other concepts (such as gravity) in order to allow the end user greater freedom in simulation parameters.
 
 
Quote
Gravity is also a meaningless concept in "Pond Mode". All you need to know is that it exists. Any change in its magnitude will only serve to change the pressure acting throughout the whole system and since the robots and the water exist in a complete balance, nothing will change at all.

 No.  Net forces acting on a body in a liquid = Force of Bouyancy - Force of Gravity = Volume of object * density of liquid * gravity - mass * gravity = gravity * (Volume * density - mass)
 
 Gravity doesn't cancel out.  Having more gravity will increase the magnitude of the resulting force.  Objects that float upwards at X meters per second on Earth will float X/6 meters per second on the moon.

Quote
It's called a SWIM BLADDER. Fish are able to manipulate their density to float up or sink down at will. Plankton do it too. As do krill and many other aquatic creatures.

 I meant in the DB universe there's nothing to construct to make you less dense.  And while we're on the subject, swim bladders have a whole set of gotchas and consequences fish have to deal with.  Among them they'd explode if they tried to rise too quickly.
 
 
Quote
It didn't need to correlate with much of the other physics. It was designed to override them.

 Override the physics system or override the physics constants?  Those are two very different things.
 
 
Quote
Frankly I have never really liked the physics of 2.4 much. That's another reason why I have still not made the switch.

 I'm sorry you feel like this, but most everyone else I've heard comment on them feels they are a step in the right direction.  The problem is that they're almost realistic, so people are better able spot discrepancies with real life.  The older physics was so absolutely unrealistic you could never really compare it with real life so its flaws were a little better masked.
 
 Do various stress test comparisons between 2.4 and 2.37.  1000 veggies with some gravity for instance.  Or several bots that tie to everything they can.  I think you'll see my point.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7725
    • View Profile
Version 2.42.6 Released
« Reply #20 on: June 17, 2006, 10:39:23 PM »
addendum:  I seperated Pond Mode for version 2.36.
« Last Edit: June 18, 2006, 02:18:50 AM by Numsgil »

Offline PurpleYouko

  • Bot God
  • *****
  • Posts: 2556
    • View Profile
Version 2.42.6 Released
« Reply #21 on: June 18, 2006, 11:51:01 AM »
Quote
This is not exactly true. Most fish are not in equilibrium density with their environment.

Swim bladders are not found in all fish. Many cartilaginous fish, such as sharks, can control their depth only by swimming; others store fats or oils for this purpose.

OK it is true that some fish such as sharks do not have bouancy control, I wouldn't go so far as to say most fish are not in equilibrium with their environment. All open water species of boney fish and a large proportion of reef fish are able to control their bouyancy. Whether they use fats and oils or swim bladders or whatever, they still have methods to control their mean internal density such that they can control their bouyancy.

Quote
At present pond mode implies only the way in which light is distributed in the sim. What if I want to run a sim with a gradient for the vegs in a total vacuum? I'm certain I mentioned this when I split corpse mode and pond mode apart into their most core features. I removed the hooks to the other concepts (such as gravity) in order to allow the end user greater freedom in simulation parameters.
If you want to simulate a light gradient in a vacuum then you could just click the nice handy toggle that disables Bouyancy in 2.37.6. You know, the one that isn't there in 2.4X
You may well have mentioned it but if you did then I was probably too busy with something else to pay much attention. I definitely don't remember it.

Quote
Gravity doesn't cancel out. Having more gravity will increase the magnitude of the resulting force. Objects that float upwards at X meters per second on Earth will float X/6 meters per second on the moon.
Yes you are right that gravity would have an effect on the magnitude[/b] of acceleration forces caused by bouyancy but for a robot that has neutral bouyancy w.r.t. its environment the net change in acceleration due to a change in gravity would be zero.

Quote
I meant in the DB universe there's nothing to construct to make you less dense. And while we're on the subject, swim bladders have a whole set of gotchas and consequences fish have to deal with. Among them they'd explode if they tried to rise too quickly.
In all 2.3 versions of DB there IS something to construct which can make a robot less dense. I put it there. It doesn't matter what it is. Maybe robots are more dense than the surroundings by default and ned to actively spend enrgy to become neutral. Doesn't matter what the mechanism is any more than it matters what mechanism is used to rotate or accelerate or fire shots. The mechanism is simpy there.
In 2.4 you just removed it.
If you want to simulate exploding because of rapid rising through the medium then fine. Do that. That wouldn't bother me. I don't really see the point but it wouldn't bother me. Just arbitrarily removing a whole chunk of pond mode is another matter though.

Quote
Override the physics system or override the physics constants? Those are two very different things.
Yes they are though from my point of view it doesn't make a bit of difference so long as the sim does what I want it to do. I need my robots to be able to control their bouyancy. That's all. they have a sysvar to set it and another to read it. in 2.4x they simply don't work. What more do I need to say?

Quote
I'm sorry you feel like this, but most everyone else I've heard comment on them feels they are a step in the right direction.
It isn't a case of not being a step in the right direction. I can see that a lot of the underlying physics are becoming more accurate. It is just that so much of the system is just "missing". I know you kind of left 2.4 as a half unfinished project and that (most of) the stuff that you actually did made those finished bits better, but the missing stuff just makes it unusable to me.

Quote
The older physics was so absolutely unrealistic you could never really compare it with real life so its flaws were a little better masked.
Yes this is true.
Possibly the entire allure of DB (at least to me)  has always been its differentness to reality. If we go 100% realistic, then how do we reconcile the way they move, turn, shoot to feed, fire ties, read each other's minds.
There really isn't much about DB that is technically realistic when it comes right down to it.
Why should the physics have to be a mock up of the reality that we know?
There are 10 kinds of people in the world
Those who understand binary.
and those who don't

:D PY :D

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7725
    • View Profile
Version 2.42.6 Released
« Reply #22 on: June 18, 2006, 01:40:51 PM »
Quote from: PurpleYouko
If you want to simulate a light gradient in a vacuum then you could just click the nice handy toggle that disables Bouyancy in 2.37.6. You know, the one that isn't there in 2.4X
You may well have mentioned it but if you did then I was probably too busy with something else to pay much attention. I definitely don't remember it.

 There's no such toggle because Bouyancy is assumed to be an emergent property of liquid properties.  To disable bouyancy you'd have to either disable gravity or turn fluid density to 0.
 
 
Quote
Yes you are right that gravity would have an effect on the magnitude of acceleration forces caused by bouyancy but for a robot that has neutral bouyancy w.r.t. its environment the net change in acceleration due to a change in gravity would be zero.

 What I'm saying is that staying at exactly neutral bouyancy is difficult, even in real life.  If you were in a sea in Jupiter and were 1% off, your acceleration would be quite different from the same scenario on Earth.  My point is that the magnitude of gravity changes the margin for error.
 
 
Quote
In all 2.3 versions of DB there IS something to construct which can make a robot less dense. I put it there. It doesn't matter what it is. Maybe robots are more dense than the surroundings by default and ned to actively spend enrgy to become neutral. Doesn't matter what the mechanism is any more than it matters what mechanism is used to rotate or accelerate or fire shots. The mechanism is simpy there.
In 2.4 you just removed it.

 I didn't remove it as much as pull the rug out from under it.  My real point is that when we reintroduce it, we should treat it as a constructable unit, and so have hooks on the cost page for changing the cost per unit, same as with shell and slime.  I'd prefer it had another use, so it wasn't just for bouyancy control, in the same way that shell isn't just for balast.
 
 A simple solution would be to change the properties of slime to serve this purpose.  Slime would either be zero volume (as it is right now) with negative mass (which makes about 0 sense in any realistic setting) or give slime some volume and 0 mass (slime has 0 mass right now).
 
 You could reason the latter as a kind of "air bubble" made from mucous.  Sort of like what lungfish create underwater.
 
 
Quote
There really isn't much about DB that is technically realistic when it comes right down to it.
Why should the physics have to be a mock up of the reality that we know?

 If you really understand the internals, it makes no difference at all.  What it comes down to is learning curve.  The easier it is for lay users to figure out what's going on in a sim, the more likely they are to pick it up and continue playing with it.
« Last Edit: June 18, 2006, 01:41:30 PM by Numsgil »

Offline Elite

  • Bot Overlord
  • ****
  • Posts: 532
    • View Profile
Version 2.42.6 Released
« Reply #23 on: June 19, 2006, 10:12:56 AM »
I'm going to have to agree with PY here. I liked buoyancy too

How hard would it be to just have .setboy increase and decrease the buoyancy of a bot? If you really want to be realistic with it you could have it control the size of an air bubble, so setting it high increases the size of the bot but keeps the mass the same, decreasing density, while decreasing it makes the air bubble smaller, increasing density. You could also use it to change the size of a bot when buoyancy is disabled.

How about that?

Removing .setboy will cause backwards compatabiltiy problems, like with PY's Alga Stratificus (which I qiute liked), so if you're aiming for complete backwards compatability for 2.4 from 2.37.6 (which Eric is) then reincluding it would be a good idea.

How a special option to 'enable buoyant pond mode', which will overide all your environmental settings to give you an environment similar to 2.37.6's pond mode with buoyancy enabled, setting y-axis gravity and fluid density automatically?

Offline PurpleYouko

  • Bot God
  • *****
  • Posts: 2556
    • View Profile
Version 2.42.6 Released
« Reply #24 on: June 19, 2006, 01:15:27 PM »
Quote
There's no such toggle because Bouyancy is assumed to be an emergent property of liquid properties. To disable bouyancy you'd have to either disable gravity or turn fluid density to 0.
Absolutely no problems with this.
That is exactly why I said that pond mode over-rides other aspects of the physics and that for the purposes of pond mode, Gravity is irrelevent as bots are assumed to have neutral bouyancy.
You are just over complicating a very simple concept.

Quote
What I'm saying is that staying at exactly neutral bouyancy is difficult, even in real life. If you were in a sea in Jupiter and were 1% off, your acceleration would be quite different from the same scenario on Earth. My point is that the magnitude of gravity changes the margin for error.
And my point is who the hell cares?
As long as the bouyancy controls work the way they are supposed to, I couldn't give a crap what real life says about it.

Quote
I didn't remove it as much as pull the rug out from under it. My real point is that when we reintroduce it, we should treat it as a constructable unit, and so have hooks on the cost page for changing the cost per unit, same as with shell and slime. I'd prefer it had another use, so it wasn't just for bouyancy control, in the same way that shell isn't just for balast.
Yes that would work. I still think this is just overcomplicating things though.

Quote
If you really understand the internals, it makes no difference at all. What it comes down to is learning curve. The easier it is for lay users to figure out what's going on in a sim, the more likely they are to pick it up and continue playing with it.
I really do not agree with you here and maybe this is why I don't really like 2.4 all that much.
Frankly I don't care about mirroring real physics so lang as the game plays well and does what you would expect it to.
The physics controls in 2.4 are over complicated and confusing. Heck if I was new to the game I would take one look at all those settings and run a mile.
Most of the time I simply can't figure out how to make the bloody things work at all. It's more guess work than anything else unless you really understand every litle detail. I much preferred a simple slider for gravity and for friction. 2.4 is impossibly over complicated. I just don't like it.
There are 10 kinds of people in the world
Those who understand binary.
and those who don't

:D PY :D

Offline PurpleYouko

  • Bot God
  • *****
  • Posts: 2556
    • View Profile
Version 2.42.6 Released
« Reply #25 on: June 19, 2006, 01:17:24 PM »
Quote
How a special option to 'enable buoyant pond mode', which will overide all your environmental settings to give you an environment similar to 2.37.6's pond mode with buoyancy enabled, setting y-axis gravity and fluid density automatically?
Good idea Elite....

Oh wait a minute...

Didn't we already have that once before?  
There are 10 kinds of people in the world
Those who understand binary.
and those who don't

:D PY :D

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7725
    • View Profile
Version 2.42.6 Released
« Reply #26 on: June 19, 2006, 03:52:36 PM »
You seem a little combative PY
 
 
Quote from: PurpleYouko
And my point is who the hell cares?
As long as the bouyancy controls work the way they are supposed to, I couldn't give a crap what real life says about it.

 The smaller the margin for error, the more precise any evolution has to be to make a bot that can both sink and swim.  It changes the fitness landscape.  It creates more seperation between any populations of sinkers and floaters.  It amplifies extremes.
 
 If you don't really care, you can just set gravity to any value you like.
 
Quote
I really do not agree with you here and maybe this is why I don't really like 2.4 all that much.
Frankly I don't care about mirroring real physics so lang as the game plays well and does what you would expect it to.
 
The physics controls in 2.4 are over complicated and confusing. Heck if I was new to the game I would take one look at all those settings and run a mile.
 
Most of the time I simply can't figure out how to make the bloody things work at all. It's more guess work than anything else unless you really understand every litle detail. I much preferred a simple slider for gravity and for friction. 2.4 is impossibly over complicated. I just don't like it.

 The problem isn't the core physics.  That's controlled by very few parameters.  The problem comes in the disconnect between the physics tab and the advanced physics tab.  Friction is also a little confusing, I'll probably combine z axis gravity and the coefficients into two sliders.
 
 Physics basically comes down to 7 choices for physics in 2.37 and 9 essential choices for physics in 2.4 (2 for friction, 2 for fluid properties, 1 for planet eaters, 1 for Zero Momentum, 1 for Brownian, 1 for bang efficiency, and 1 for gravity).
 
 Probably the physics tab needs to dissappear and have the advanced physics controls replace it.
 
 However, I think you're exagerating to say that physics made more sense in 2.37.  How many people had any idea what swimming factor did?  You had to look in the source code to figure it out.
 
 I think you're being a little neophobic.  You were intimately familiar with the older version's physics and code, and you find it discomforting to find yourself essentially a newbie again in some areas.
 
 -------------------------------
 
Quote
How hard would it be to just have .setboy increase and decrease the buoyancy of a bot?

 My only concern is in the "automatic" nature of the old version.  Shell decreases bouyancy, yet its effects are effortlessly removed by setting bouyancy to 0.  Ideally I'd like to see some consequences that have to be weighed.  Having setboy be a air bubble or something similar, as you describe, so that it effects final bouyancy instead of overriding it, would be acceptable to me.  I'd also like to see some costs involved in increasing bouyancy the same way cost is involved in decreasing it (constructing shell).
 
 
Quote
How a special option to 'enable buoyant pond mode', which will overide all your environmental settings to give you an environment similar to 2.37.6's pond mode with buoyancy enabled, setting y-axis gravity and fluid density automatically?

 Which brings up a new point/idea:  Several preset default settings you can select.  Old school pond mode, F1 default, etc. that automatically set various options to the proper value for you.

Offline PurpleYouko

  • Bot God
  • *****
  • Posts: 2556
    • View Profile
Version 2.42.6 Released
« Reply #27 on: June 20, 2006, 09:43:10 AM »
Quote
You seem a little combative PY  
Sorry bout that. I've been spending too much time debating at EVC Forums

Quote
The problem isn't the core physics. That's controlled by very few parameters.
You are right about that. The problem really has very little to do with the actual physics engine. it is all about how intuitive the controls are. I just don't like them in 2.4. I find the controls have little or no meaning until I go into the advanced tab. then they are a littl more like the old stuff.

Quote
Probably the physics tab needs to dissappear and have the advanced physics controls replace it.
I would definately prefer that.

Quote
However, I think you're exagerating to say that physics made more sense in 2.37. How many people had any idea what swimming factor did? You had to look in the source code to figure it out.
It didn't really do anything much at all. Supposedly it enabled wiggly things to provide locomotion by wiggling but somewhere down the line they  lost the ability to do so. I could never really get my head round Carlo's code that made them do it so I never really managed to re-introduce it.

Quote
I think you're being a little neophobic. You were intimately familiar with the older version's physics and code, and you find it discomforting to find yourself essentially a newbie again in some areas.
I won't argue with this.

Quote
My only concern is in the "automatic" nature of the old version. Shell decreases bouyancy, yet its effects are effortlessly removed by setting bouyancy to 0.
Shell was never part of the bouyancy equation. That wasn't by design. It should have been part of it. Increase in shell should undoubtedly cause negative bouyancy which would have to be countered by raising the value in .setboy. Maintaining a non-zero setboy value should in turn cost energy proportional to the difference between it and zero. that too was never implemented but probably should be.
You just have to think of .setboy as a universal all inclusive method of modifying bouyancy. It's one of those things that has no correlation in reality. Bit like shots, ties and so on. It doesn't need a correlation. if we force bots to be realistic then we have to redesign the entire game from the ground up.

Quote
Which brings up a new point/idea: Several preset default settings you can select. Old school pond mode, F1 default, etc. that automatically set various options to the proper value for you.
That would work just fine provided that the new physics is able to generate a world that mimics traditional pond mode.

I don't know whether the physics are just buggy but I can't make them do anything much at all. Sometimes I change a setting then change it back and it refuses to go back at all. often the only way to get it back is to turn of DB and restart it. I seem to recal Testlund reporting something similar a while back.
There are 10 kinds of people in the world
Those who understand binary.
and those who don't

:D PY :D