Poll

What do you like doing most in RA2/RA3/robot combat games?

Building robots
Local single player battles with manually-controlled robots
Local multiplayer battles  with manually-controlled robots (PvP with controllers or a shared keyboard)
AI-only tournament battles
Other?  Please comment below.

Author Topic: Robot Rumble 2.0 - Robot Combat Simulator - Under Development  (Read 29716 times)

Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #580 on: May 24, 2019, 08:54:57 AM »
Played it: here is the feedback
1. Flipper motors are still broken, they flip and break off
2. Music is added back, and i cant mute it
3. Worst one, you either get one hit KO'ed or do a one Hit KO

1. and 3. - I've been testing with Holy Wings versus a horizontal spinner.  I have found that if you leave everything on Holy Wings at 1 mm UHMW, the heavy spinning bar will indeed one-hit destroy any UHMW that it touches.  This is fairly realistic.  1 mm UHMW is like protecting your robot with cardboard against a big spinner.  If you armor everything with about 5 mm of steel, however, Holy Wings survives a lot longer. 

The biggest problem I am finding now is that the motors don't have enough torque to lift the flipper arms fast enough to work as flippers against another heavyweight.  Now that all of the UHMW has been replaced by thicker steel armor, Holy Wings is just a wedge that can wave "hi!" with its flipper arms.  I'm not sure how the bigger AmpFlow PMDC motors would work as flippers in real life either.  To my knowledge people tend to use pneumatics for heavyweight flippers.

2. So sorry about the music!  I'm not sure what happened here, but we definitely need to fix this. 

Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #581 on: May 24, 2019, 09:04:25 AM »
Okay so I just tried this out and just from the first few minutes I have this feedback:

  • Manta's flipper is really havok-y. It manages one or two flips that yeet me right into the ceiling, and then it spends the rest of the fight propelling itself around the arena at warp-speed while its flipper panel glitches into another dimension. I've had the same issue with Goodbye Kiss's axe in the past, dunno if that helps at all. Earthquake's flipper doesn't close properly, preventing it from firing more than once or twice. TR3's flipper seems to work fine.
  • None of my robots' weapons work anymore. I get that you can now choose which button fires them, but the dropdown just says "Button1", "Button2" etc, with no clue as to what those buttons actually are. Is it possible that you added controller compatibility for Bugglebots, but broke keyboard compatibility in the process?
  • Bots still seem too fragile. I thought this might have been just Fait Accompli and its 3mm steel, which is the bare minimum for a real-life HW, but then I noticed that sometimes, I'll start taking massive damage while just driving normally, without my opponent anywhere near me. I can only imagine that my robot is somehow rubbing against the floor, and the game is interpreting that as damage. Ignoring the fact that a 6WD robot like Fait Accompli shouldn't be touching the floor with anything but its wheels in the first place, this should not be enough to kill you in two seconds.

1. Re: Manta - Oof.  I struggled with Manta for about three hours on Tuesday night.  In the game, pneumatics use a plunger system that pushes on the lifting arm to cause it to raise.  In Manta's case, the plunger is really far back, which means that the physics engine can do wonky things as the plunger penetrates the flipper arm and the physics engine computes the impulse required to resolve the collision.  I thought I had it working well, but you aren't the first one to report that it is broken.

Back to the drawing (tweaking) board...

2. Re: Unworking weapons - There is still a bug that I haven't found that sometimes prevents button inputs from being read.  This is an intermittent problem.  My workaround has been to go back to the Robot Selection screen and start over.  For me starting over fixes it about 70% of the time.  I definitely need to track this down, but each time I think I have solved it, it pops up again.

3. Re: Driving damage - You are absolutely right.  Driving normally shouldn't damage a robot.  Somehow or another something is registering a collision with a ton of impulse.  Would you mind sending me Fait Accompli?  I would like to do some testing.  I have a few things I have in mind to improve driving dynamics that might solve this problem as well.

Offline F1Krazy

  • Team VVV - Veni Vidi Vici
  • Heavyweight
  • Posts: 730
  • Rep: 4
  • Nothing is easy the first time round
    • View Profile
    • Awards
  • See profile for gamer tags: Yes
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #582 on: May 24, 2019, 09:37:55 AM »
1. Re: Manta - Oof.  I struggled with Manta for about three hours on Tuesday night.  In the game, pneumatics use a plunger system that pushes on the lifting arm to cause it to raise.  In Manta's case, the plunger is really far back, which means that the physics engine can do wonky things as the plunger penetrates the flipper arm and the physics engine computes the impulse required to resolve the collision.  I thought I had it working well, but you aren't the first one to report that it is broken.

Back to the drawing (tweaking) board...
I think Manta's flipper has been broken for a while tbh, I managed to make it glitch out at least once in the previous build.

2. Re: Unworking weapons - There is still a bug that I haven't found that sometimes prevents button inputs from being read.  This is an intermittent problem.  My workaround has been to go back to the Robot Selection screen and start over.  For me starting over fixes it about 70% of the time.  I definitely need to track this down, but each time I think I have solved it, it pops up again.
I'll try that. I assume Button1 is Space, then?

3. Re: Driving damage - You are absolutely right.  Driving normally shouldn't damage a robot.  Somehow or another something is registering a collision with a ton of impulse.  Would you mind sending me Fait Accompli?  I would like to do some testing.  I have a few things I have in mind to improve driving dynamics that might solve this problem as well.
Yep, I'll send it over right away.
My DSL Showcase


Offline kix

  • Hey boss!
  • *
  • Posts: 2573
  • Rep: 11
  • We need my rep to go higher
    • View Profile
    • Awards
  • Discord: Username#8552
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #583 on: May 24, 2019, 10:05:19 AM »
1. and 3. - I've been testing with Holy Wings versus a horizontal spinner.  I have found that if you leave everything on Holy Wings at 1 mm UHMW, the heavy spinning bar will indeed one-hit destroy any UHMW that it touches.  This is fairly realistic.  1 mm UHMW is like protecting your robot with cardboard against a big spinner.  If you armor everything with about 5 mm of steel, however, Holy Wings survives a lot longer. 
Problem is that the bots i was fighting had Steel armour. This OHKO also tends to happen to prebuilt bots

The biggest problem I am finding now is that the motors don't have enough torque to lift the flipper arms fast enough to work as flippers against another heavyweight.  Now that all of the UHMW has been replaced by thicker steel armor,
Make Servo motors that have high torque?
As for the flipper arms, i have been using Hinged actuators which are op

Holy Wings is just a wedge that can wave "hi!" with its flipper arms.  I'm not sure how the bigger AmpFlow PMDC motors would work as flippers in real life either.  To my knowledge people tend to use pneumatics for heavyweight flippers.
Well tbh that is old wings, it does not use a hinged actuator for flipping

Another oddity that i find is that when keys are binded, it keeps the control (lets say space), but if i rebind it again, the old imput key (space) will stay written while the new key will be functional

Also Metals have this odd Blue-ish tint where tint RGB is set on all to 1

AAAAAAlso like it how lightweight bots have great driving, but HW bots tend to bounce, 4wd ones atleast
H
EEEEEEEEEEEEEEEEE

Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #584 on: May 24, 2019, 10:56:21 AM »

I think Manta's flipper has been broken for a while tbh, I managed to make it glitch out at least once in the previous build.

True.  Its flipper geometry makes it tough for the physics engine.  I might have to do some cheating with the visuals to get rid of the problem entirely.

I'll try that. I assume Button1 is Space, then?

Correct.  This is reconfigurable, but space is the default.

Yep, I'll send it over right away.

Excellent!  I'll see what I can do.  I had a brainstorm last night which will take some work but might make driving a lot smoother for all robots.

Offline F1Krazy

  • Team VVV - Veni Vidi Vici
  • Heavyweight
  • Posts: 730
  • Rep: 4
  • Nothing is easy the first time round
    • View Profile
    • Awards
  • See profile for gamer tags: Yes
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #585 on: May 24, 2019, 11:02:41 AM »
Did my PM get through to you? The first time I tried, I got a "there's already a file with this name" error, and the second time I tried, I got a "you've already sent this PM" error. And GTM annoyingly doesn't save messages to your outbox by default, so I don't know whether either of them actually got sent.
My DSL Showcase


Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #586 on: May 24, 2019, 11:07:09 AM »
Problem is that the bots i was fighting had Steel armour. This OHKO also tends to happen to prebuilt bots

Was this happening even when the chassis had steel armor?  Steel armor on a large body should give a ton of HP that should take a while to chip through. 

The new puncturing system might be the culprit.  If a robot hits a UHMW piece that is easy to puncture, it will generate a ton of puncture damage that will carry over into the chassis.  I've seen cases where a spinner punctures a UHMW or MDF piece for a 45x damage multiplier.  Instead of just saying that the component is dead and stopping there, that damage is then transferred to the chassis.  This could easily disable a bot.  Need to nerf this.

The prebuilt bots don't use the same system, so we can give them infinite HP without affecting the Botlab robots.

Make Servo motors that have high torque?
As for the flipper arms, i have been using Hinged actuators which are op
Well tbh that is old wings, it does not use a hinged actuator for flipping


Ah!  I was wondering how your flippers were so powerful!  @tashic put them in, but I haven't experimented with them yet.



Another oddity that i find is that when keys are binded, it keeps the control (lets say space), but if i rebind it again, the old imput key (space) will stay written while the new key will be functional

So weird...


Also Metals have this odd Blue-ish tint where tint RGB is set on all to 1


Lighting probably?  Either that or post-processing.  They both could be contributing to an excess of blue.  I'm pretty sure the texture is grayscale.


AAAAAAlso like it how lightweight bots have great driving, but HW bots tend to bounce, 4wd ones atleast

Driving again!  Need to fix this! :)  I have something in mind.  I don't know if it will work, but it has the potential to make driving a lot smoother.

Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #587 on: May 24, 2019, 11:15:37 AM »
Did my PM get through to you? The first time I tried, I got a "there's already a file with this name" error, and the second time I tried, I got a "you've already sent this PM" error. And GTM annoyingly doesn't save messages to your outbox by default, so I don't know whether either of them actually got sent.

I have not received a PM from you.  I just sent you a PM with my email address.

Offline kix

  • Hey boss!
  • *
  • Posts: 2573
  • Rep: 11
  • We need my rep to go higher
    • View Profile
    • Awards
  • Discord: Username#8552
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #588 on: May 24, 2019, 11:23:44 AM »
Lighting probably?  Either that or post-processing.  They both could be contributing to an excess of blue.  I'm pretty sure the texture is grayscale.
Its throughout the whole game, so it might be post processing, UHMW is white tho
Also any chances to use an option to disable post processing? I have a feeling that that might be the culprit to the lag on lower end machines

As for driving, Just remove friction for now
H
EEEEEEEEEEEEEEEEE

Offline Guldenflame

  • This is very inconsistent
  • Posts: 1090
  • Rep: 3
  • The third most forgettable person here.
    • https://www.facebook.com/
  • Awards Donated money for hosting 2019 BOTM Winner
    • View Profile
    • Everything is too OP
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #589 on: May 24, 2019, 12:30:28 PM »
My current hypothesis on the whole Instant KO thing is that everything is treated like a part of the chassis.  Most of the time when I lose a bot, it's because it got hit in the wheel, weapon, or a thin weapon support.

I guess at the moment, everyone has RA2 logic in their head where if it isn't a part of the main chassis, it won't damage the bot when hit.

Offline Hoppin

  • E-Girl
  • *
  • Posts: 1806
  • Rep: 11
  • Sket - 1000% more toxic than TheRoboteerm
  • Awards BOTM Winner
    • View Profile
    • Awards
  • Discord: Hoppin#0013
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #590 on: May 24, 2019, 12:32:34 PM »
My current hypothesis on the whole Instant KO thing is that everything is treated like a part of the chassis.  Most of the time when I lose a bot, it's because it got hit in the wheel or a thin weapon support.

I guess at the moment, everyone has RA2 logic in their head where if it isn't a part of the main chassis, it won't damage the bot when hit.

My assumption is that it being all chained to the chassis ripples in. Killing the chassis
Things I did & done

Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #591 on: May 24, 2019, 12:45:29 PM »
My current hypothesis on the whole Instant KO thing is that everything is treated like a part of the chassis.  Most of the time when I lose a bot, it's because it got hit in the wheel or a thin weapon support.

I guess at the moment, everyone has RA2 logic in their head where if it isn't a part of the main chassis, it won't damage the bot when hit.

My assumption is that it being all chained to the chassis ripples in. Killing the chassis

Agreed.  The easy fix is to stop all damage propagation, but getting parts to break off nicely becomes an issue in this case.

Another alternative is to just roll all components that are contained within the same rigid body into a single, combined HP number.  Don’t do breakage at all for components within a rigid body.  Everything that isn’t connected via a hinge joint would then be part of the chassis.

I’ll give this a go today to see how it feels.

Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #592 on: May 24, 2019, 12:55:28 PM »
Lighting probably?  Either that or post-processing.  They both could be contributing to an excess of blue.  I'm pretty sure the texture is grayscale.
Its throughout the whole game, so it might be post processing, UHMW is white tho
Also any chances to use an option to disable post processing? I have a feeling that that might be the culprit to the lag on lower end machines

As for driving, Just remove friction for now

Interesting.  We’ll see if we can add disabling post processing to the “Low Quality” setting.

Friction coefficient is currently set to zero for all chassis materials.  I believe the reason things drive a little weirdly is that when the chassis skids along the floor the collision resolution impulse direction isn’t always purely vertical.  I have had luck in the past adding sphere colliders to the bottom of a robot to fix this.  I might try to do this automatically for all botlab robots.  I can see where this might make wedges a little less effective, as the tip of the wedge would be just slightly off the ground.  Maybe it wouldn’t be too bad if all robots were subject to the same limitation.

Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #593 on: May 24, 2019, 01:45:09 PM »
Did my PM get through to you?

I just tried out Fait Accompli and fixed the concussion damage issue when the robot is driving around.  It turns out that I had set the minimum impulse limit ridiculously low such that pretty much every heavyweight would damage itself just by driving.

I also noticed that the axe didn't move when triggered.  I tried dropping the thickness of the steel used.  That did the trick.  Here are my settings:

Axe head (pyramid) - 6 mm steel - 0.481 kg
Axe shaft (cube) - 3 mm steel - 0.637 kg

Now to see what it does to other robots...

Offline F1Krazy

  • Team VVV - Veni Vidi Vici
  • Heavyweight
  • Posts: 730
  • Rep: 4
  • Nothing is easy the first time round
    • View Profile
    • Awards
  • See profile for gamer tags: Yes
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #594 on: May 24, 2019, 01:55:36 PM »
Did my PM get through to you?

I just tried out Fait Accompli and fixed the concussion damage issue when the robot is driving around.  It turns out that I had set the minimum impulse limit ridiculously low such that pretty much every heavyweight would damage itself just by driving.

I also noticed that the axe didn't move when triggered.  I tried dropping the thickness of the steel used.  That did the trick.  Here are my settings:

Axe head (pyramid) - 6 mm steel - 0.481 kg
Axe shaft (cube) - 3 mm steel - 0.637 kg

Now to see what it does to other robots...
Oh, so the weapon actually did work, it was just too heavy for the motor? Well gee, now I feel stupid  :facepalm: I did actually run another test with Defcon and its weapon worked just fine (it even gyros now!), so I guess that was all that was wrong. I'll try your suggested weights, and also see what happens if I put a counterweight on the other end.

Glad you've got the driving issue sorted out, though!
My DSL Showcase


Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #595 on: May 24, 2019, 01:59:57 PM »
I'll try your suggested weights, and also see what happens if I put a counterweight on the other end.

Glad you've got the driving issue sorted out, though!

Thanks!

I also have a sneaking suspicion that when I tell the physics system to give me a certain number of Newton-meters of torque, the system isn't giving me that much torque.  I'll have to set up a test rig in Unity to see for sure.  I feel like the motor you are using should be able to swing that axe head around in real life.

FWIW I have a 6" diameter motor in the works for weapons, similar to the old E-teks.  I'm not sure how it would work for an axebot, but it should be killer for spinners.

Offline F1Krazy

  • Team VVV - Veni Vidi Vici
  • Heavyweight
  • Posts: 730
  • Rep: 4
  • Nothing is easy the first time round
    • View Profile
    • Awards
  • See profile for gamer tags: Yes
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #596 on: May 24, 2019, 02:17:58 PM »
I'll try your suggested weights, and also see what happens if I put a counterweight on the other end.

Glad you've got the driving issue sorted out, though!

Thanks!

I also have a sneaking suspicion that when I tell the physics system to give me a certain number of Newton-meters of torque, the system isn't giving me that much torque.  I'll have to set up a test rig in Unity to see for sure.  I feel like the motor you are using should be able to swing that axe head around in real life.

FWIW I have a 6" diameter motor in the works for weapons, similar to the old E-teks.  I'm not sure how it would work for an axebot, but it should be killer for spinners.
I lowered the weight of the axe and it still didn't swing in the test area... then I switched the motor's activation button to Button 2 and then back to Button 1, and it worked perfectly. I think what might actually be happening is, where I've ported my bots over from the previous version, there's some sort of compatibility issue and it doesn't know what key should activate the robot's weapon until I go into the bot lab and tell it.

It now also doesn't self-right anymore, which might be because of the lowered weight, or it might be a torque issue. It was just about able to self-right in the previous build.
My DSL Showcase


Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #597 on: May 24, 2019, 02:34:50 PM »
I feel like the motor isn't quite torquey enough, and doesn't generate the tip speed it should.

For the moment I have set the minimum tip speed for rotating puncturing objects to be 10 m/s.  The axe isn't quite there, and doesn't seem to be able to get above 10 m/s there no matter how much I drop the mass.  The axe itself is breaking off from being hit before the tip does any damage to its opponent.

My brain is full and I need to head home.  I'll think about it on the drive home tonight.

Some options:
1. Reduce the tip speed requirement.  This is bad, because all of a sudden just bumping into stuff will cause damage.

2. Increase the motor torque.

Thinking about #2, I just checked the torque and saw that the stall torque on an AmpFlow A40-300 is only 27 Newton-meters.  This is pretty low, considering how we are asking it to swing a big heavy axe.

I'm going to pop in the geared version of the motor.  I have an Ampflow A28-400G already modeled.  Hopefully the 220 Newton-meters of the geared version will be enough to swing the axe in a satisfying way.

EDIT - Now that I'm thinking about it, even a 220 Nm (approximately 22 kg-meter) gear motor might have trouble turning over a 100 kg robot with an arm like that.  Need to do some calculations to be sure, but...

EDIT OF MY EDIT - Considering that the existing motor can just barely swing the arm now, it is no wonder that tip speed is not reaching 10 m/s.  The tip is basically just falling 1 meter from its peak, with barely any help from the motor.  A 1 meter fall only gives a speed of 4.4 m/s.  Must have bigger motor with more torque.

Offline Guldenflame

  • This is very inconsistent
  • Posts: 1090
  • Rep: 3
  • The third most forgettable person here.
    • https://www.facebook.com/
  • Awards Donated money for hosting 2019 BOTM Winner
    • View Profile
    • Everything is too OP
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #598 on: May 24, 2019, 05:24:32 PM »
Another fun discovery: attacking pieces that have fallen off a robot still deal damage.

Offline cjbruce

  • Middleweight
  • Posts: 364
  • Rep: 10
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #599 on: May 24, 2019, 05:30:14 PM »
Another fun discovery: attacking pieces that have fallen off a robot still deal damage.

To the robot the piece fell off from?!  That is completely unintended and hilarious. :)