Jump to content

[1.3] Kerbal Joint Reinforcement v3.3.3 7/24/17


ferram4

Recommended Posts

Alright, version 1.4 is up, which should hopefully fix most of the wobble people were having between parts of very different sizes. If not, there's always the config.xml to play with to see if that helps.

Link to comment
Share on other sites

Bugreport:

After docking:

NullReferenceException: Object reference not set to an instance of an object

at KerbalJointReinforcement.KerbalJointReinforcement.UpdatePartJoint (.Part p) [0x00000] in <filename unknown>:0

at KerbalJointReinforcement.KerbalJointReinforcement.FixedUpdate () [0x00000] in <filename unknown>:0

(Filename: Line: -1)

NullReferenceException: Object reference not set to an instance of an object

at KerbalJointReinforcement.KerbalJointReinforcement.UpdatePartJoint (.Part p) [0x00000] in <filename unknown>:0

at KerbalJointReinforcement.KerbalJointReinforcement.FixedUpdate () [0x00000] in <filename unknown>:0

(Filename: Line: -1)

..

..

..

spamming the console and slowing the game ;)

Link to comment
Share on other sites

A.g. pointed that issue out to me shortly after I released v1.4; the v1.4.1 hotfix is up to fix that error, as well as an error where the max force a joint could apply was not handled properly.

Link to comment
Share on other sites

Just installed the 1.4.1, i was working on a new engine block for a super-heavy tanker, with 1.3 it was turning itself into a black hole and then kessler bomb due to its own gravity (or due to the kraken), now it stays together :D Yay 488t engine block in orbit, with almost no strut... I'll test radial stuff and my Laythe jet again tomorrow, now time for some sleep.

Here it is:

KtL16Rm.jpg

I had a weird failure with 1.3 on the launch pad, i'll try again tomorrow if i can reproduce it with 1.4.1, since it seems the radial attachments are fixed now...

Link to comment
Share on other sites

I've still managed to summon the joint-Kraken with some of those radial attachments, so I don't consider it solved by any means. Go for a radial attachment with a lot of mass on the end of a long moment arm and you'll see it happen.

Link to comment
Share on other sites

Awesome, thanks for the update! All we need now is the plugin to make the rocket one part to even further reduce the part count sorta like the weild plugin and ksp will be almost lag free till 0.23 =)!

this would be a VERY bad idea.

I'll not go into detail, it has been explained before.

Link to comment
Share on other sites

this would be a VERY bad idea.

I'll not go into detail, it has been explained before.

Too bad. I mean more for Space stations really. Because making huge ones creates so much lag. I might have to just remake my current one now that i have this plugin installed.

Link to comment
Share on other sites

Too bad. I mean more for Space stations really. Because making huge ones creates so much lag. I might have to just remake my current one now that i have this plugin installed.

Why remake? The plugin acts on existing vessels as well, no need to destroy and re-create the station.

Link to comment
Share on other sites

Awesome, thanks for the update! All we need now is the plugin to make the rocket one part to even further reduce the part count sorta like the weild plugin and ksp will be almost lag free till 0.23 =)!

Such exists, but you can't make the entire rocket one part.

Link to comment
Share on other sites

I've still managed to summon the joint-Kraken with some of those radial attachments, so I don't consider it solved by any means. Go for a radial attachment with a lot of mass on the end of a long moment arm and you'll see it happen.

I've seen that reported as a problem with stock as well somewhere in the support forum, and could reproduce it myself, so it might be some problem with the physics engine itself. One thing I noticed when observing it in slow motion at that time was that the wobble appeared to start with some of the heavy tanks twisting around the truss used to attach them, so the root of the problem may be in the angular motion. The Dynamic Warp mod is useful for observing how things happen frame by frame. :)

Link to comment
Share on other sites

Two issues to report: 1, and I keep forgetting to say this so you may have fixed something else to undo this, but: time acceleration of 10 million in a medium Kerbin orbit leads to an explosion on going off rails when KJR is enabled. (RSS, obviously; replicate by setting max warp in the cfg to 10m). I have it set to 6 now which should be safe, so it's not a big issue.

2. Per the Saturn IB I just shows shots of the in RSS thread, I rescaled the Mk1 pod to 4m width, and made a copy of the 2.5m heatshield rescaled to 4m width. Every time I launch with KJR, there's a structural failure between the heat shield and the pod; when I revert to non-KJR the rocket holds together! Weird.

I'll post the stuff you'll need to replicate on the RSS thread.

Link to comment
Share on other sites

@a.g.: It's a known issue with working with discrete-time simulations. Basically, if the timestep isn't small enough the error in the simulation can add energy faster than the damper in the joint can remove it. The problem is that making the damper too strong can lead to things shaking themselves apart due to floating point errors.

@NathanKell: Issue 1 sounds weird... how quickly does the game end up scaling down from 10 million, is it instantaneous or does it ramp? It might just be that physics is activating before the timewarp is "completed" and so things are breaking like they did in stock KSP a few versions ago. Does first coming down to a slower speed, then dropping out of timewarp fix it?

With Issue 2, I'm gonna guess two other things are going on here: one is that you've got stretchy tanks installed for that Saturn IB; it seems like the huge masses and sizes of the stretchy tank parts causes lots of problems with whatever they're connected to. The second thing is that the launch clamps jump on load when this happens; it seems like RSS exacerbates a current issue with the clamps where they shift a little bit on load; it's somewhat intermittent though, so I don't know exactly what causes the shift to occur.

Link to comment
Share on other sites

1. Nope, that was what we tried: go down in stages, using the built-in lerping between warps. The collisions were weird--it said collided with launchpad, despite being in orbit!

2. See the latest post in the RSS thread for the craft file, apologies for delay. I tried it first without struts between pod and heatshield, then with. Finally, removing the struts again, I get about a 25% success rate on load to pad, and then another failure chance on first stage event. Note that with the struts it's quite solid without KJR, which is very strange; that's why I mention the issue. I know ST has issues, but they're usually worse on stock.

I absolutely don't use launch clamps; I know better. :]

Link to comment
Share on other sites

Awesome work, this is fried gold for my 3.75 rockets.

One question on linear vs angular joint limits:

When exactly does each come into play?

i.e. Is linear exclusive to node-to-node connections whereas angular is to radial connections?

Also on node-to-node connections:

Does 1 connection == 1 joint?

Link to comment
Share on other sites

Linear is movement, angular is rotation. And yes, except for decoupler and launch clamp reinforcement, it just replaces stock connection joints with custom configured ones.

Thanks, so now in theory I should be able to increase linBounce which will allow joints to survive the occasional failure caused by rapid oscillations during unpack/physics load.

Now what about oscillations in flight? You would think to increase linDamp but the description says it equates to strength? Also paradoxically, the description for linSpring says it equates to stiffness. o_O

Link to comment
Share on other sites

Anyone tried undocking with the latest version?? Just added a bit to my main station and cant undock my salvage ship. Further testing and I cant seem to undock any docking ports once connected. Only thing the log shows is a single NullRefEx each time 'undock' button in context menu is selected.

I am using FusTek cbm if that matters.

Link to comment
Share on other sites

@a.g., Ferram: i posted a stock craft that exhibits the "infinite-expanding-wobble-then-kraken" issue a little while ago, i'll try it with the new version to see how it has evolved.

Also:

NsrfXOg.jpgP7fRfiH.jpg

This one krakenized itself on the launch pad, before i could do anything. Then the core (about 250t itself, not really light) went up floating in the breeze, like a balloon !! That was both weird and hilarious. I think i found another way to break the physics engine :D

@KhaosCorp: you are victim of the undocking bug. It's not related to this plugin, but to the stock game itself. I really hope the devs can fix it in v0.23.. in the mean time the only way is to undock whatever main section contains the faulty dock, separate it, then re-dock what you need redocked. Or edit your save file by hand with a rather complicated procedure (especially if you have many parts).

Edited by Surefoot
Link to comment
Share on other sites

@NathanKell: Alright, I'll look into both. No doubt I've caused something stupid to happen.

@egreeSS: linSpring controls the stiffness of the spring that tries to bring the parts back together

linDamp controls the strength of the damper that tries to counteract any movement caused by the spring / outside forces

linLimit controls how much wiggle room there is before these particular values kick in

linBounce determines whether the joint will "bounce" off the limit; so if it hits that limit and linBounce = 1, a force will be applied through the joint so that it will instantly go back in the other direction. Don't increase this, seriously.

@KhaosCorp: What does the output_log.txt say is the source of the exception? I know that there is a similar issue in stock KSP where docking ports will not undock.

@Surefoot: STOP BREAKING MY MOD! :P

What exactly happened? Did it just start to jiggle around and then popped upwards? What's the output_log.txt say, if anything?

Link to comment
Share on other sites

@Surefoot: STOP BREAKING MY MOD! :P

What exactly happened? Did it just start to jiggle around and then popped upwards? What's the output_log.txt say, if anything?

:D haha sorry i was too busy laughing out that i didnt check the logs, to be honest i'm not even sure it's related to your mod (for the floating part). And yeah it started to jiggle by itself, then all the parts warped inside out (like there was a sudden black hole in the middle) and it kinda threw the central part up, which then started to float like a balloon. No, i dont take drugs, i promise !

It's a bit late for me now, i'll see tomorrow if i have time to reduce the problem to a simple mod-less design.

(edit) the "proof of concept" radial mount orange tank is still wobbling itself to death on the launchpad (it's random, sometimes just warping a bit helps). Radially mounted stuff creates weird interactions... My current design is an attempt to avoid the kraken, by using a very heavy, huge core based on 5m parts, but it seems i just opened a new can of worms :P

Edited by Surefoot
Link to comment
Share on other sites

@KhaosCorp: you are victim of the undocking bug. It's not related to this plugin, but to the stock game itself. I really hope the devs can fix it in v0.23.. in the mean time the only way is to undock whatever main section contains the faulty dock, separate it, then re-dock what you need redocked. Or edit your save file by hand with a rather complicated procedure (especially if you have many parts).

Oh that's so lame..this thing was supposed to be temp....now its kinda messin up my whole station!...gonna have to rip it off I think.

Would be nice if Squad put stuff like this in the 'Known Issues' thread...

Link to comment
Share on other sites

Just loggin in to say thank you. I've wanted something like this for a LONG time. I never understood why two metal tanks, welded together to withstand rocket-engine stress levels... would wobble like a table that's missing a leg. THANK YOU. XD

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...