Show Posts

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 - CyarSkirata

Pages: [1] 2 3 4 5 6
1
I'm suddenly wondering about an idea for things like hubmotors and axle bots. Would it potentially be viable to have alternative versions of motors and gearboxes where the game considers the axle to be the root part, and the main body the axle?

So you place the gearbox on the robot by its axle, then place the motor on the end of the gearbox.

2
For this next build I have combined spinners into a single damageable object.  This means that the whole spinner will break off together.

My Tombclone will break off Circumvolution's spinner in a single hit every single time.  It makes for really short fights.

Ah i really love this thing you said, maybe you should enable an option to disable breaking stuff off if on an axle

Just out of curiosity. Does the multi part spinner get combined HP as a single component, or does it have the HP of the base part?

It adds up the total health of the spinner components and consolidates them into a single number.  Right now health = mass * (health per mass of the material).  More mass = more health.

Circumvolution Spinner = 250 health
Tombclone Spinner = 500 health

Tombclone can do thousands of Joules of damage on a single hit, so Circumvolution's spinner is destroyed every single time.

Tbh irl Hori would beat a vert easily. If you are uncomfortable with it you can always add a small HP multiplier
Else lower the damage slider to let the fight last longer

I mean, that isn't universal, IRL it depends on the individual spinners and how they meet. In my experience, the only inherent horizontal advantage is more from the way their manoeuvres change where on the weapon's path you're meeting it, since that lets them hit the side of the vert almost every time regardless of the vert's efforts to avoid it.

3
Spinners in Feb28 have something weird going on with their mass or something. My verts start to wheely when their weapons are spinning.

It's by far at its most noticeable with S3. Drive forward with the weapon off, it's fine. Drive forward with the weapon on, and it's more overhead thwackbot than vert.
  [ Quoting of attachment images from other messages is not allowed ]

Is it the reaction torque that occurs when the weapon starts up?

Circumvolution is hard to drive too.  It tends to flip on its back a lot.

Nothing about acceleration. You can leave it to spin all the way up and it still flips over.

4
Spinners in Feb28 have something weird going on with their mass or something. My verts start to wheely when their weapons are spinning.

It's by far at its most noticeable with S3. Drive forward with the weapon off, it's fine. Drive forward with the weapon on, and it's more overhead thwackbot than vert.
 

5
PS, Cyar, I’m tracking down the improper spinner breakage bugs in Circumvolution.  I think I have fixed them all.  Now that they are fixed a new problem has arisen.  When three out of the four quarters break off the weapon becomes so unbalanced that it teleports all over the arena.  I have a few potential fixes in.  We’ll see.

Hmm. Hopefully something does work.

6
I've been thinking of composing some music loops for the game now that I have a bit of extra storage space for Cubase projects. Is there anything in particular you'd like?

7
I just thought of a theory regarding one of the issues I've been seeing with flywheels.

Perhaps damage is now being calculated based on the size or weight of the individual part that scores the hit? That would explain why I see bars doing just fine but all the flywheels I have do basically nothing, since said flywheels all have their teeth attached as separate components rather than included as part of the main body of the weapon.

8
I even built a completely new robot to the same concept with fewer parts (such as using a hollow cylinder for the flywheel) and that one's even worse. I've seen its wheels break seemingly just from the weight of the robot pressing down on them, and the flywheel consistently just comes clean off after it scores one or occasionally two of its pathetic hits.

 :vista:

The only damage that should exist in the game right now is due to spinner hits.  A robot pressing down on its wheels shouldn't do any damage.  Do you think its weapon is damaging its own wheels?

Also, just curious, but what damage slider setting are you using?

The first tests were on 25. I then turned it up to 100 in the hope that maybe the opposing bots would start taking damage - since Circumvolution and its new sibling would struggle to get much more fragile - but nothing seemed to actually change.

9
Okay, so... I've just tried the feb16 build, and Circumvolution's gotten catastrophically worse for reasons I can't figure out. The entire robot disintegrates at the slightest provocation, and the weapon is simply incapable of damaging anything more than exposed wheels during its brief existence. If anything, the shrapnel is probably more dangerous than the robot itself.

I even built a completely new robot to the same concept with fewer parts (such as using a hollow cylinder for the flywheel) and that one's even worse. I've seen its wheels break seemingly just from the weight of the robot pressing down on them, and the flywheel consistently just comes clean off after it scores one or occasionally two of its pathetic hits.

 :vista:

10
I've been having problems with Circumvolution's flywheel again. This is the exact same version of the bot that you already have so you should be able to test it. I'll also send an updated version when I've put the new motors in.

Frequently, damage caused to a part on the flywheel is dealt to one of the parts further back in the parent chain.
That parent part promptly comes off after taking only the tiniest fraction of the damage its HP should be able to absorb.
When this happens, that parent part rarely takes more than a few of its child parts with it, and never takes all of them.

Typically this manifests as a tooth hitting an opponent, one of the parts of that tooth taking the last bit of damage needed to finish it off, and at that point that entire quarter of the weapon comes off whole along with the heavy steel-50 cylinder that serves as the root for the entire weapon, while the other three quarters of the weapon remain attached despite now missing their shared parent part.

11
I'm not seeing anything too weird here.  The COMs are in the correct location when the wheels break off and the rigidbody masses are set to 0.00001 kg like they are supposed to be.  Is it possible that the behavior that you are seeing is actually the correct result of the angular momentum of the big, fast vertical spinner?

The issue isn't with getting gyro. It's with getting something that resembles gyro, but the robot will - for example - end up with one side completely hovering while it drives back and forth in a straight line, or take a vert hit and be sent spinning through the air but only be capable of spinning on the axis of the axle that's lost its wheel.

12
Would someone mind posting their absolutely worst case robot showing the following problem:

1. A wheel breaks off.
2. The robot tilts unnaturally.

None of the robots I have built seem to have the problem.  I have tried to replicate the issue without success, but have a few changes I would like to try.

My clusterbot has developed this bug sometime since the last time I tested it.
The wheels are pretty exposed, so I suggest putting one of them in the arena under AI control and using Circumvolution (reach, side-wedges and good bite for getting at the sides) to pull the wheels off while it tries to fight back, to test the bug under combat conditions, where the target bot is moving pretty much all the time.

 
 

13
Panic Attack's issues with having too much traction are fixed, and so is the srimech torque. The version with the slow un-wheelfixed weapons still struggles to steer, but the fast version is back in business. :)

Sadly, no improvement in having the thwackbot be able to thwack in the arena.

14
I've been having another go at getting my thwackbot to work, and I feel like part of the problem is a lack of traction with the arena floor.

It uses the stock 12.5 inch wheels, and in the workshop test area it thwacks so hard it bounces the wheels up off the ground, but out in the arena it needs a several metre run-up to turn over at all, and when it does do so it's too sluggish to have any real impact even if it did hit.

15

I didn't lose track of it though. :/
When I reworked it to add the beefed up axle, I even re-attached *everything* to the axle.

Also, on that note, it's impossible to break off the decorative parts of Panic Attack's srimech.
As in: when the main plate of the srimech breaks off, everything decorative stays attached.

Gotcha.  The problem is that an axle is not currently defined as its own component.  It doesn’t have health, at least not by itself.  It can’t be hit or take damage. Rather, it is part of a motor.  In order to break things off of the axle, they must have their own collider.  Since the collider is removed from anything marked as decorative, the decorative pieces attached directly to the axle won’t come off until the motor itself is destroyed.

Need to think about this...

Oh. When I said axle, that's not what I meant. I meant that I added a beefed up steel cylinder to the spinner, and made every other section of the weapon be child parts to that.

It's entirely possible for that cylinder - the parent part to the entire weapon - to come off from the teeth hitting opponents. On top of that, when that parent part does come off, it very rarely takes its child parts with it.

As for Panic Attack:
The srimech has three aluminium components attached to the main slab. Each of those parts has decorative parts attached to it.
When those break off, the decorative parts attached to them go with them. When the main slab breaks off, the decorative parts attached to it don't go with it. They remain attached to the gearbox.
The skirts also take their decorative parts with them when they break off.

Having talked about the issues with the two bots together, I do now have a theory. Perhaps the fault lies specifically in the components that are attached directly to the axle? Since that's the case both for the main plate of Panic Attack's srimech, and the cylinder in Circumvolution's weapon.

16
Circumvolution has had an occasional problem for awhile where when a part of the flywheel breaks off, child components of that part will stay attached, hovering and still a capable part of the weapon.

Sometimes, the central axle will come off from the impacts of the teeth on an opponent (already concerning), and only take one of the flywheel sections with it.

This is a result of how the components are attached to each other in the tree structure.

Each component has a parent.  If a parent breaks off, all of the children also break off.  If a component's parent has not broken off, the component and all of its children remain.

In Circumvolution's case, this means that there are many many bits and pieces attached to other bits and pieces.  It is easy to lose track of which piece is parented to which when building.

I am looking to do some automatic collider reduction.  This should help reduce the problem, but won't eliminate it entirely.  Fewer colliders = things break off in larger chunks.  It would still be possible for something to break off in an unrealistic way if components were attached to components in a weird way.

I didn't lose track of it though. :/
When I reworked it to add the beefed up axle, I even re-attached *everything* to the axle.

Also, on that note, it's impossible to break off the decorative parts of Panic Attack's srimech.
As in: when the main plate of the srimech breaks off, everything decorative stays attached.

17
Circumvolution has had an occasional problem for awhile where when a part of the flywheel breaks off, child components of that part will stay attached, hovering and still a capable part of the weapon.

Sometimes, the central axle will come off from the impacts of the teeth on an opponent (already concerning), and only take one of the flywheel sections with it.

18
Is there a place folks are uploading their RR2 creations for others to download?

Not yet, as far as I'm aware. But I do know there are one or two ideas in the works.

19
Is the robot taking significantly longer than this to complete a turn for you?

Yup, longer than that.

In other news, I just built a thwackbot that works perfectly in the test area, but I can't make it thwack in the arena. :(

20
Panic Attack turns fairly quickly with the keyboard if you don't try to drive forward or backward at the same time as you turn.

Okay... Now I'm confused.
I have the exact opposite problem. I can only get it to turn at a decent rate by getting up some speed and drifting it.

What is your setting for Wheel Mass Ratio?  I've been using 10%.

Mine's at 10 too

Pages: [1] 2 3 4 5 6