This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Messages - cjbruce
Pages: 1 ... 12 13 14 15 16 17 18 [19] 20 21 22 23 24 25 26 ... 49
361
« on: January 08, 2020, 07:49:00 AM »
-Driving seems to be lower and more sluggish than the previous builds
Now that the overall robot mass is kept constant, all robots should maintain their inertia. We got use to the inertia dropping once the spinner got up to speed. Maybe this explains the sluggishness?
it kinda does, but if i gear up the motor (or down), in theory, the bot should drive faster, but in game the bot bounces
I just found (and fixed) a contributor to the problem. My previous approach to compensating for the loss of spinner weight was to add an extra gravitational force at the location of the spinner. I forgot to remove this when I added the Makeup Mass to the chassis. The net result of having both is that the gravitational force on the spinner was twice as strong. In my "Tombclone"-style spinner this meant that the spinner was pulled down hard, and the wheels weren't getting any grip. Driving was incredibly sluggish. With the extra gravity force removed the driving went back to normal. Depending on the location and weight of the spinner, this could have either a small or large effect on the driving characteristics of a given robot.
362
« on: January 08, 2020, 06:04:29 AM »
Honestly the hits is kinda feeble to me.
And the sparks look nice but a bit weird
Buffing the kickback! :) I have some good spark footage I need to look at again on my phone. Just had the kids do some grinding with the angle grinder -- I'm thinking of trying to match that...
363
« on: January 08, 2020, 06:02:55 AM »
So basically all of my bots have lost their gyro with this new build, even the ones with huge spinners. And it was working fine in the Jan 01 build
True! The purpose of these recent builds is to see what happens when the mass of spinning objects is reduced to almost nothing. This is true in pretty much every other game. I think you are right -- having gyro is a good thing. Maintaining mass at 100% does create all sorts of instability though, so it is a balancing act to drop the amount of mass until the problems go away.
364
« on: January 08, 2020, 05:37:03 AM »
Ok so tested the new build.
The good:
-Sparks actually are great.
I'm glad you like them! They need some tweaking, but I thought they weren't bad for a first attempt. The bad:
-Gyro is non existant
The slider should help with that. I'm thinking that we put in a slider that goes from 0% - 50%. In the 01January2020 build, I had minimum spinner masses of around 30%, depending on RPM. In the current build my minimum spinner masses are around 1%. -Driving seems to be lower and more sluggish than the previous builds
Now that the overall robot mass is kept constant, all robots should maintain their inertia. We got use to the inertia dropping once the spinner got up to speed. Maybe this explains the sluggishness? -The hits seem to be mixed bag. My hits seem to be okay, i can hit the bot and the weapon will not slow down immensly. Some bots fail to do that
The hit response depends on bite. If bite is small, then a glancing blow occurs and there isn't much kick. If bite is big, then a "good hit" occurs and the spinner will stop instantly and create a large amount of kick (impulse). -When a disc/bar spinner hits the ground it loses all the torque and the bot doesnt even react properly, its like the bar weighs 0 kilos. Craplectric ant self right like that anymore, it just tumbles until it self rights, even tho the spinner is running at 150+ mph when it hits the ground.
I'm planning to increase the kick on glancing blows. This should help with the self-righting.
365
« on: January 08, 2020, 05:16:46 AM »
Are you thinking what I'm thinking? 'cause I'm thinking its SLIDER TIME!  I'm putting in a slider to control the maximum amount of spinner mass reduction. I'm also thinking I need to increase the amount of kickback that occurs on a spinner glancing blow. That should help a with the self-righting.
366
« on: January 07, 2020, 07:20:05 PM »
The 07January2020 Alpha Build is up on itch.io! https://robot-rumble.itch.io/buildsI'm really excited about this one. Please let me know how driving and spinner hits feel!
367
« on: January 07, 2020, 03:43:33 PM »
New build coming soon! [Restored] Spinner collision impulse response is now back to 50% of the value it had in the late-December builds. In those builds, all of the spinner's angular momentum was converted into linear momentum via P = L / radius. Converting only 50% of the angular momentum into linear momentum feels about right to me for "good hits". Please let me know if hits feel good! [Added] Added sparks for spinner hits against steel. 
368
« on: January 07, 2020, 06:08:54 AM »
Here's what I'm seeing for both "Panic Attack" and "Panic Attack WF". I can't really tell a difference between the two versions. Both versions were capable of lifting the 100 kg crate with both the lifting forks and the srimechs. 
369
« on: January 07, 2020, 05:36:31 AM »
Yup. Limits are set. Both do the exact same 90 degree arc, but the non-wheelfixed one runs better, while being kinda flimsy.
Make sure all four weapon gearboxes on the non-wheelfixed one are set to 10/1, to match the wheelfixed one, then put each of them in the test area and give the srimech a waggle.
If your simulation is running roughly the same as mine, then on the jan 6 build you should see the non-wheelfixed version twitch faintly (reasonably good), while the wheelfixed one pivots clear into the air (practically kills the robot in battle.)
Shoot. The work I am doing right now shouldn't affect the way these robots are handled, nor should it treat them any differently. All I have been doing for the past two weeks has been spinner stuff. I will try fiddling with the two robots as you have suggested to see if I can find any differences in the way they are reconstructed for combat... I just put up the 07January2020 bleeding edge build. Would you mind taking a look at the driving again for your horizontal and vertical spinners? http://www.robot-rumble.com/bleedingedgebuilds/For this build, I added the makeup mass to the chassis. The overall mass of the robot shouldn't change even when the spinner RPM increases and its mass drops. This should mean that the robot's inertia and weight should remain the same even when the spinner is spinning.
370
« on: January 06, 2020, 08:45:57 PM »
While I'm here, I might as well also mention: This test does actually fix the suicidal torque when I use the srimech on non-wheelfixed versions of Panic Attack. Unfortunately it also makes it too flimsy.
The issue still remains though with it being effectively unable to steer when I gear the weapons low without the wheelfix.
The problem there being that with the wheelfix, they go completely limp when put under any minor shock from moving parts of opponents. With both gearboxes on each weapon set to 10/1 - meaning a total ratio of 100/1, the effective limit of a single gearbox - they move fast enough that it isn't an issue in battle.
But the wheelfix makes them unuseable at the more authentic speed given by setting each gearbox at 10/0.4 - a speed at which the torque would have little enough of an effect that I'd be able to use the srimech effectively.
So right now the choice with the bot is between being nearly incapable of steering, and being nearly incapable of self-righting. 
Do you have limits set on the srimech motor? The wheelfix shouldn't actually do anything if that is the case. If the code sees that limits are set, it should treat it exactly the same as if a wheel is attached. I just checked both versions of Panic Attack you sent, and neither of them have a Spinner Mass Reducer -- both have limits set on the motor. Both versions should behave exactly the same.
371
« on: January 06, 2020, 08:37:57 PM »
Just put up a new 06January2020 Bleeding Edge Build:
http://www.robot-rumble.com/bleedingedgebuilds/
The purpose of this build is to test driving with your big spinners. I reduced the mass of all big spinners by about 90% from where they were this should significantly improve the stability of the simulation for big heavy spinners.
In particular, if you have a shell spinner let me know how it drives with this build!
Okay. So... First off, I imagine you already know this, but this test does make spinners nearly incapable of scoring effective hits.
You are correct. I removed spinner hit impulse response from this build so that it would be easier to isolate driving problems. I was finding that for Mental Breakdown the spinner had an annoying habit of bumping into the ground and making the robot really hard to drive. I will put the spinner hit impulse response back in once we are happy with the driving feel. My verts handle so well they're actually harder to aim because of the oversteer. Circumvolution in particular is pretty ridiculous, it almost feels like it steers faster than without the weapon, though that's probably just a psychological thing from me over-correcting.
My more standard horizontals basically feel like they don't have a weapon at all, in all aspects of handling, despite one of them having a "32 kg" flywheel with the torque to spin up in under 3 seconds.
My shell and my ring both have basically the same problem; they feel like they've already lost their weapons, and their opponents don't notice their hits at all.
This feedback is incredibly helpful! Thank you! I think I might have just solved the problem for good. I am adding mass to the chassis in the original location of the spinner. The amount of mass added is exactly equal to the amount of mass lost as the spinner gains speed. The idea is to avoid the gyroscopic instability by reducing the the mass of the spinner, but to maintain the weight and inertia of the robot overall. Thank you thank you for testing with your robots! New build is on its way tomorrow morning!
372
« on: January 06, 2020, 08:00:45 AM »
Just put up a new 06January2020 Bleeding Edge Build: http://www.robot-rumble.com/bleedingedgebuilds/The purpose of this build is to test driving with your big spinners. I reduced the mass of all big spinners by about 90% from where they were this should significantly improve the stability of the simulation for big heavy spinners. In particular, if you have a shell spinner let me know how it drives with this build!
373
« on: January 04, 2020, 09:07:39 PM »
Would you mind sending me S3? I would like to get the spinning-related quirks nailed out.
Sure. [ Quoting of attachment images from other messages is not allowed ]
WOWSERS! I haven't seen that problem before. It seems like the center of mass of the robot is far off to the side for some reason. Like maybe the whole robot got shifted sideways? I don't know what the fix is. It doesn't have anything to do with a blur cylinder. Maybe some numbers need to be tweaked in the .RR2Bot file? @tashic, any ideas?
374
« on: January 04, 2020, 06:43:19 PM »
That’s kind of what I was thinking too. I also would like to take into account ramming and being launched up high, then crashing into the ground.
Spinners are definitely the special case. They have their own special collider. I was hoping to do everything else without adding anything special.
375
« on: January 04, 2020, 01:31:22 PM »
Back at it working on spinner impulse response...
I have broken things down into two categories: glancing blows and direct hits.
Glancing blows should slow down the spinner, push the robots apart, and shoot sparks (if appropriate).
Direct hits should put as much kinetic energy into crushing something as possible. If the spinner has enough kinetic energy to break something, it will, slowing the spinner in the process. If the spinner doesn’t have enough kinetic energy, it will come to a complete stop. Reaction impulse will push the spinner backward and the target forward.
At some point I’m going to have to deal with the fact that hammers are going to have less than 1% of the kinetic energy of a spinner. Damage will need to work differently. I suppose that is a problem for the future though.
376
« on: January 04, 2020, 08:22:16 AM »
I noticed this glitch in the 03January build:
[ Quoting of attachment images from other messages is not allowed ]
EDIT: The robot that is gyrodancing has no spinning weapons.
I've had that problem on S3 ever since the first build with part damage. Except on S3, it's the bot's default state of being. Regardless of damage, drive controls, and whether the weapon is running.
Would you mind sending me S3? I would like to get the spinning-related quirks nailed out.
377
« on: January 03, 2020, 07:02:36 PM »
I’m pulling out the center of mass calculations that I did this afternoon. I will have a clean version up tomorrow morning.
Chassis damage will need to wait a bit.
Will that fix clipping thru issues and random teleportation?
Shoot! I just realized that the probable reason for the glitches is the fact that the blur cylinder isn’t disabled every hit like it was in previous builds. Prior to this every hit the speed of the spinner was dropped to zero. Now the speed only drops a little on most hits. Thus the physics impulse is applied to the colliding objects continuously. This will take some tweaking to fix. Maybe a few back and forth iterations until we are satisfied. We’ll see...
378
« on: January 03, 2020, 05:09:11 PM »
I’m pulling out the center of mass calculations that I did this afternoon. I will have a clean version up tomorrow morning.
Chassis damage will need to wait a bit.
379
« on: January 03, 2020, 04:32:35 PM »
I noticed this glitch in the 03January build:
EDIT: The robot that is gyrodancing has no spinning weapons.
Oh no. I forgot to take wheels into account in the recalculation of center of mass when stuff gets broken off.
Pages: 1 ... 12 13 14 15 16 17 18 [19] 20 21 22 23 24 25 26 ... 49
|
Show Posts