Offcenter Perms

[kill343gs]

My first decent looking DSL bot has a problem. A big one.

When I tested it after building it it worked fine. As soon as I entered an exhibition the attachment points on the perm 132s were offcenter. The bot acts like any other unstable VS, moving around wildly and almost uncontrollable. I have tested it several times and the problem is the same, Entering an exhibition match causes the perms to glitch. If I rebuild the bot and test it in the garage it works fine again, as soon as I enter the event its glitched again. Does anyone have any idea what the problem could be? The perms are mounted off of those FBS side plate things if that could have anything to do with it.

[Madiaba]

Hey, K-3. Nice to see you around.
I've seen this in some of Turret applications based upon a dual-servo set up. Can you send me the bot file and I'll look at it.

[Jules]

This has also happened to my 3 or 4 times. Just last night infact.
My motor was mounted directly to the baseplate, so I don't think where it's connected matters.
After trying to re-enter matches, etc. like you did, I had to remove the perm and rebuild everything that was on it.

[Clickbeetle]

It has something to do with skirts and skirt hinges.  Those are just glitchy components even after I totally remade them for the 2.1 patch.  My bot Darkling did that glitch, and I found if I removed the front skirt hinge it got better.  Try using skirts not on a hinge, and/or non-skirt components on a hinge.

[kill343gs]

If that's the case I think its more the hinges than the skirts itself. The skirts are basically just thing wedges/armor, hinges are burst motors with no power. Things that are supposed to move that can't are probably more likely to glitch than a stationary component.

Anyways, I rebuilt the bot and it works fine now, so no big deal.

[Madiaba]

If that's the case I think its more the hinges than the skirts itself. The skirts are basically just thing wedges/armor, hinges are burst motors with no power. Things that are supposed to move that can't are probably more likely to glitch than a stationary component.

[/COLOR]

Anyways, I rebuilt the bot and it works fine now, so no big deal.

Exactly, that's what I have found. Probably because of the huge amount of consistent data-processing needed for calculating hinges(i.e.*GMID_HAVOK_HINGE_CONSTRAINT), they are prone to glitching. Especially if:
-there are a great many hinges
-there are several proximal hinges
-there are sequential hinges.
 
Besides consistently calculating the spatial correlative parameters between body/axle, Havok must also factor in (1)simulated 'tolerances/play/slop' in hinges, (2)mass/weight, (3)inertia, and (4)angular momentum.
On the other hand this combination of factors is what allows for 'Glitch' assembly techniques.
 
When I was tweaking out dual-servo cannon aiming for MadBeaver's Tanks, I ran into this problem. When I attached one servo to another, the simulated tolerances/slop would cause the Host's axle to sag severely, ruining any attempt at controlled aiming. Since we don't have access to this coding, I had no choice but to create new servos that were lighter. So I went through all of the spin, servo and burst motors to find the most stable body/axle combination, and then both shrank and lightened them (body/axle) up. This greatly helped the sagging problem, but soon I had an attachment problem, because the lightened axle would not support something attached to it; but it would 'fly off' of the bot, and I could not fix it in the shop; and thus had to discard the whole bot and rebuild a new one.
 
ACAMS and I had a couple discussions trying to figure out how to get around this. In the end there was a dual hinge (servo) set up that works well, but is very temperamental during assembly. This is why I haven't released this dual 'lateral' and 'elevation' cannon aiming system that I used on his tanks and several other 'Combat' vehicles. It works well, but assembly could cause manual baldness.
 
So K-3, I see this as a similar problem of calculating hinges. I'm not sure of the 'axle' weight on the hinges, but I'll bet it is light, and would at times have a propensity to fly off the body when something is attached to it.
-----------
 
If interested one can see their beta and final User and AI operation here (see "Self Propelled-4" vids): https://gametechmods.com/forums/showthread.php?t=841
 
.

[kill343gs]

This would also explain why the skirts on my original bot sank through the floor.