lo-fi Posted August 11, 2016 Author Share Posted August 11, 2016 @Shadowmage lastest commit fixes rigidbody offset from wheel bug. This was a badly needed revision, so good catch! Link to comment Share on other sites More sharing options...
V8jester Posted August 11, 2016 Share Posted August 11, 2016 This is not 1.1.3. This is still 1.0.5 so don't anyone go flipping out. A Little something to keep the momentum going we all really appreciate all the work that goes into KF Link to comment Share on other sites More sharing options...
lo-fi Posted August 11, 2016 Author Share Posted August 11, 2016 Thanks for sharing, that's awesome! Really like the HAB 6062 video too. Link to comment Share on other sites More sharing options...
Shadowmage Posted August 11, 2016 Share Posted August 11, 2016 3 hours ago, lo-fi said: @Shadowmage lastest commit fixes rigidbody offset from wheel bug. This was a badly needed revision, so good catch! Good to hear I'll take another crack at it tonight with your linked project + latest commits. Should know more by tomorrow; seems like it shouldn't be hard at all to get the friction working. Link to comment Share on other sites More sharing options...
lo-fi Posted August 11, 2016 Author Share Posted August 11, 2016 Great I'm working on getting the angular joint drive working so we have steering while sticky friction is engaged. Link to comment Share on other sites More sharing options...
V8jester Posted August 11, 2016 Share Posted August 11, 2016 57 minutes ago, lo-fi said: Thanks for sharing, that's awesome! Really like the HAB 6062 video too. Oh man I really need to revive that project... My cousin was going to help me with that one as a lot of time would get wasted on charicter animations. But RL intervened, and I just don't have a clue where to take that story. Hopefully someday that one will get some life breathed back into it. RL keeps me stupid busy (building engineer) Link to comment Share on other sites More sharing options...
Shadowmage Posted August 11, 2016 Share Posted August 11, 2016 2 hours ago, lo-fi said: Great I'm working on getting the angular joint drive working so we have steering while sticky friction is engaged. One thing I noticed in the code... The YDrive is supposed to be used for the suspension, correct? However, it is 'disabled' when I examined the joints that the script created. I believe it needs its 'mode' set to JointDrive.POSITION in order to actually enable the spring/damper for the YDrive. (I could be wrong though... joints and I don't exactly get along very well) https://github.com/shadowmage45/KSPWheel/blob/1ac805591ad803d7c7ec4d92a6b28d37fd7cf76c/VSProject/KSPWheel/Component/SLWheelCollider.cs#L312-L316 Link to comment Share on other sites More sharing options...
lo-fi Posted August 11, 2016 Author Share Posted August 11, 2016 IIRC, that's a deprecated setting from U4. But that's only a vague recollection. What Unity version are you running? Link to comment Share on other sites More sharing options...
Shadowmage Posted August 11, 2016 Share Posted August 11, 2016 2 minutes ago, lo-fi said: IIRC, that's a deprecated setting from U4. But that's only a vague recollection. What Unity version are you running? 5.2.4 (as-per the setup for KSP modding for 1.1). Could be part of the problems I was seeing last night (I even tried a multi-rigidbody setup with each wheel its own rigidbody and it was still very wonky). However, I probably should upgrade to 5.4 if/when it is available... might be some joint-related fixes in there, haven't looked too hard (other than to verify that they did -not- fix the wheels properly). Link to comment Share on other sites More sharing options...
lo-fi Posted August 11, 2016 Author Share Posted August 11, 2016 Hehe, the problems you were having last night were definitely my fault. I'm developing in 5.3, which can either be seen as a mistake, or forward thinking for 1.2.. Drive mode deprecated in 5.3: http://answers.unity3d.com/questions/1120193/jointdrivemode-is-obsoleted-on-unity-53-so-what-no.html I doubt we'll be ahead of 1.2, so we can either continue with that, or I can rewrite for U5.3 What do you reckon? Link to comment Share on other sites More sharing options...
Mycroft Posted August 11, 2016 Share Posted August 11, 2016 Excuse a nobody for butting in, but if I understand you correctly, I think you're saying it won't beat 1.2 to release. If this is the case, perhaps it would be best to aim for 1.2? IMHO. Feel free to disregard it. Link to comment Share on other sites More sharing options...
Shadowmage Posted August 11, 2016 Share Posted August 11, 2016 I would say we should probably use 5.3 / 5.4 and develop for the future (e.g. KSP 1.2). Sounds like it has entered the QA phase, so it is likely less than a month out from release. Will likely take us about that long to get all the wheel quirks sorted out and get everything integrated (though would be nice to be able to start testing KSP integration before that; not sure if that would be possible with the differences in Unity versions). I'll setup a 5.3/5.4 Unity install tonight and begin using that for the wheels development and testing; sounds likely that the version differences were part of what I was seeing last night. Link to comment Share on other sites More sharing options...
lo-fi Posted August 11, 2016 Author Share Posted August 11, 2016 Indeed, it would be nice to get stuff tested in game... Who do we talk to about getting early modder access to 1.2? I'd hope we qualify! Otherwise, recompiling for 5.2 shouldn't be a big deal. I ported an incredibly complex 4.2.2 joint thing I made a while back to 5.3 in minutes, so it's unlikely to cause us much work. Can't remember why I landed on 5.3 - I messed that up somewhat. Link to comment Share on other sites More sharing options...
Shadowmage Posted August 11, 2016 Share Posted August 11, 2016 Initial testing with my existing test model, using Unity 5.3.6 + your latest commits, shows much more consistent and expected results Cleaning things up a bit and hooking in the friction model now, will likely have more information on that a bit later this evening. One problem that I'm noticing is that the joints are applying the suspension forces along the axis of the suspension rather than along the axis of the hit.normal. We might have to do some cheating to dynamically 'fix' that problem. The friction model can help with minor inequalities in the suspension setup, but may need to do some cheating to align the springs to the hit.normal axis somehow... something to think on. Link to comment Share on other sites More sharing options...
lo-fi Posted August 11, 2016 Author Share Posted August 11, 2016 Great to hear! At least it's working for you now. I mused about that too.... Not sure that's fixable within the joint configuration itself - I've tried a lot! Link to comment Share on other sites More sharing options...
OrbitalBuzzsaw Posted August 12, 2016 Share Posted August 12, 2016 (edited) Did I read that right? It's WORKING!!!!?! Edited August 12, 2016 by OrbitalBuzzsaw Link to comment Share on other sites More sharing options...
Mycroft Posted August 12, 2016 Share Posted August 12, 2016 3 hours ago, lo-fi said: Who do we talk to about getting early modder access to 1.2? I'd hope we qualify! If you downloaded the game through Steam, early access is available. Or at least that’s what I heard. Link to comment Share on other sites More sharing options...
Shadowmage Posted August 12, 2016 Share Posted August 12, 2016 41 minutes ago, lo-fi said: Great to hear! At least it's working for you now. I mused about that too.... Not sure that's fixable within the joint configuration itself - I've tried a lot! Yeah, don't think there is anything within the joint config that would help either. My initial thoughts were to dynamically reposition/re-anchor the joint so that it were always facing along the hit.normal in addition to dynamic spring-rate adjustment for the incidence angle between the suspension and hit.normal... but it might be a bit weird to get that figured out. My alternate thought, after seeing that the spring rates appear to make sense in newtons-meter units, would be to calculate the off-axis force the spring is adding and to negate that manually through another force in the opposite direction. Likely won't be perfectly accurate, but may be good enough for our purposes. Going to have to work on getting it fixed next though as it is interfering with the friction dynamics. I also had to disable the constraints you had setup for lateral grip and steering as they were also interfering with the force-based friction model. The lateral grip constraint might be used for sticky-friction (which it worked well for), but we'd need to find a way to enable/disable it only when needed (and setup the z-drive for longitudinal sticky friction as well). I don't think the steering constraint was needed at all though, as the actual turning response from steering arises dynamically from the application of friction to the angled wheels (which applies a rotating force around the parent body) (unless there was an intended use that I missed/am unaware of). All told though, it is going well so far (on the friction model integration). Mostly it was just plugging the update code back in and replacing a few variables that had gone missing. I now have the vehicles accelerating, steering, and braking again; though the errant feedback from angled suspension is making it hard to test the friction code much (e.g. vehicle rolls a bit during a turn... which results in even more velocity towards the outside of the turn, or during acceleration the vehicle noses up a bit... which results in negative acceleration from the suspension). Have not noticed any jittering even when I was cranking up the damper settings, but will have to do more testing on that with some multi-rigidbody vehicle setups with varying mass ratios. So.. will spend a bit of time toying with fixing the suspension-force-axis problems next, as it needs to be fixed before much more testing can be done on the suspension stuff. Link to comment Share on other sites More sharing options...
lo-fi Posted August 12, 2016 Author Share Posted August 12, 2016 3 minutes ago, Shadowmage said: My alternate thought, after seeing that the spring rates appear to make sense in newtons-meter units, would be to calculate the off-axis force the spring is adding and to negate that manually through another force in the opposite direction. Likely won't be perfectly accurate, but may be good enough for our purposes. Given what we're doing, "good enough" will definitely suffice! 4 minutes ago, Shadowmage said: but we'd need to find a way to enable/disable it only when needed That's easy, just set the constraints to "free" or turn the spring value down. 8 minutes ago, Shadowmage said: (e.g. vehicle rolls a bit during a turn... which results in even more velocity towards the outside of the turn, or during acceleration the vehicle noses up a bit... which results in negative acceleration from the suspension) Can you not account for that with a scalar product of the suspension axis against the hit normal? You get a nice 1 when they're aligned and a big fat 0 when at perpendicular! 20 minutes ago, OrbitalBuzzsaw said: Did I read that right? It's WORKING!!!!?! Early days. We've gone through the stages of "is this even a tractable problem for mere mortals?", "here are some ideas", "here are some ideas that kinda work in isolation", and now we're at "hmmmm, with a bit of tweaking this might actually work". Don't buy a ticket for the hype train just yet 18 minutes ago, Mycroft said: If you downloaded the game through Steam, early access is available. Or at least that’s what I heard. Good call. That was true for the initial 1.1 release, but it wasn't that early and I'm not sure if it still applies. I'll check. Link to comment Share on other sites More sharing options...
Shadowmage Posted August 12, 2016 Share Posted August 12, 2016 16 hours ago, lo-fi said: Given what we're doing, "good enough" will definitely suffice! That's easy, just set the constraints to "free" or turn the spring value down. Can you not account for that with a scalar product of the suspension axis against the hit normal? You get a nice 1 when they're aligned and a big fat 0 when at perpendicular! Good call. That was true for the initial 1.1 release, but it wasn't that early and I'm not sure if it still applies. I'll check. Didn't get too much more time to work on it last night after my last update. Will be spending a bit more time this evening on cleaning up the suspension force application -- as you say a dot product of the hit.normal and suspension axis (or inverse dot?) should give me the force-correction factor that I'm looking for. With any luck that will let me test the friction code integration a bit better. From there it should just be a matter of tweaking the friction curves and figuring out when to enable/disable the sticky friction constraints. Re: pre-release -- I didn't see any pre-release option in steam for 1.2, and I doubt that SQUAD will be doing a public pre-release this time given the harsh feedback they were subject to last time from non-steam users. Would certainly be nice to start testing it in-game sooner rather than later... but later might be our only option in this case. Link to comment Share on other sites More sharing options...
lo-fi Posted August 13, 2016 Author Share Posted August 13, 2016 Cool, sounds like we're on the same page I'm enquiring about getting special access like the testing team. Would you be happy to sign an NDA with squad the same as the testing team do, should they be amenable? Might come to nothing, but worth a try. I'm getting quite excited about reinventing the wheel again... I think I'm aboard the KF hype train. Co-driving it, I suppose Looking further forwards, there's a load of stuff I haven't released... If anyone is interested in some modeling and texturing for half finished crazy ideas, do shout. @V8jester, if you're still game, I'd love to get that roll cage IVA going once all the existing stuff is working again. There's also the trailer hitch, magnetic loading arm, bulldozer blade and an idea for a canadarm that will rock the KSP world's socks off if I can make it work. And an awesome modular rover that @SpannerMonkey(smce) came up with... So much stuff... Link to comment Share on other sites More sharing options...
Madrias Posted August 13, 2016 Share Posted August 13, 2016 I'm actually starting to get excited here. Link to comment Share on other sites More sharing options...
SpannerMonkey(smce) Posted August 13, 2016 Share Posted August 13, 2016 (edited) 4 hours ago, lo-fi said: Cool, sounds like we're on the same page I'm enquiring about getting special access like the testing team. Would you be happy to sign an NDA with squad the same as the testing team do, should they be amenable? Might come to nothing, but worth a try. I'm getting quite excited about reinventing the wheel again... I think I'm aboard the KF hype train. Co-driving it, I suppose Looking further forwards, there's a load of stuff I haven't released... If anyone is interested in some modeling and texturing for half finished crazy ideas, do shout. @V8jester, if you're still game, I'd love to get that roll cage IVA going once all the existing stuff is working again. There's also the trailer hitch, magnetic loading arm, bulldozer blade and an idea for a canadarm that will rock the KSP world's socks off if I can make it work. And an awesome modular rover that @SpannerMonkey(smce) came up with... So much stuff... Hiya, I've still got all the parts and files for that beasty, so more than willing to update and tweak for usability (chances are the several things about it that can be improved, it may even have a decent paint job) , As before always willing to pitch to make it all come together. Will the hypnodrive be returning? Spoiler Spoiler Edited August 13, 2016 by SpannerMonkey(smce) Link to comment Share on other sites More sharing options...
Shadowmage Posted August 13, 2016 Share Posted August 13, 2016 5 hours ago, lo-fi said: Cool, sounds like we're on the same page I'm enquiring about getting special access like the testing team. Would you be happy to sign an NDA with squad the same as the testing team do, should they be amenable? Might come to nothing, but worth a try. I'm getting quite excited about reinventing the wheel again... I think I'm aboard the KF hype train. Co-driving it, I suppose Looking further forwards, there's a load of stuff I haven't released... If anyone is interested in some modeling and texturing for half finished crazy ideas, do shout. @V8jester, if you're still game, I'd love to get that roll cage IVA going once all the existing stuff is working again. There's also the trailer hitch, magnetic loading arm, bulldozer blade and an idea for a canadarm that will rock the KSP world's socks off if I can make it work. And an awesome modular rover that @SpannerMonkey(smce) came up with... So much stuff... I've got no problem with an NDA for testing purposes... it seems to be the standard course for closed beta testing on games, which I've done quite a few of over the years. Please let me know what you find out Link to comment Share on other sites More sharing options...
V8jester Posted August 13, 2016 Share Posted August 13, 2016 5 hours ago, lo-fi said: ..... @V8jester, if you're still game, I'd love to get that roll cage IVA going once all the existing stuff is working again. ...... Absolutely. I think you had the latest and greatest build on that thing. And did you ever get around to reshaping around the Kerbal hands? That was really the only "issue" if you could call it that. Yeah that little cage was really cool! Link to comment Share on other sites More sharing options...
Recommended Posts