[1.1.3] Surface Experiment Pack - EVA science for KIS/KAS (v1.4.2 6/30/2016) *Now with less explosions*

[vardicd]

6 hours ago, CobaltWolf said:

does it affect similar KAS/KIS parts? Like the KAS port connector fuel line thing... Whatever its called.

I don't have any other KIS/KAS parts unlocked in my career game as yet to test, but i'll boot up a sandbox later to test when I get time to play, but I can say I've not seen this behavior previously. I've been using the same version of KAS/KIS, and surface experiments for quite some time, and they are all the most recent versions. I've set up surface experiments before with no issue, which is why it confused the hell out of me that this just randomly started happening, however I did narrow it down a bit, my crash seems to be coming from the newer power port parts, the 4 plug power pole and the new 2 port power plug module, which I've never used before. These fellas:

However if I use the older power port, it works just fine no crashy:

Edited August 1, 2016 by vardicd

[Ruedii]

OK, after a heated conversation with the KIS devs, due to some issues with understanding whether the issue was a bug or a needed feature.   (I've found not to question matters of pride and call something a feature request when devs claim it's not a bug.  Makes them feel all warm and fuzzy being needed, instead of feeling like having made some sort of mistake.  Some dev's can't grasp the definition of bug-upstream workaround needed as not being their fault.) It seems the workaround for the ground issue is to adjust the mount node to have a little more offset from the collide.

For the plugs, you just need to cutoff their colider to not include the prongs.  (Simple enough).  However, I would recommend instead simply having the prong mounts connect with the pipes in question directly instead. This is probably a better option, although it requires more changes.

Edited August 2, 2016 by Ruedii
Details

[CobaltWolf]

On 8/1/2016 at 6:43 AM, vardicd said:

I don't have any other KIS/KAS parts unlocked in my career game as yet to test, but i'll boot up a sandbox later to test when I get time to play, but I can say I've not seen this behavior previously. I've been using the same version of KAS/KIS, and surface experiments for quite some time, and they are all the most recent versions. I've set up surface experiments before with no issue, which is why it confused the hell out of me that this just randomly started happening, however I did narrow it down a bit, my crash seems to be coming from the newer power port parts, the 4 plug power pole and the new 2 port power plug module, which I've never used before. These fellas:

Huh. They're set up the same in Unity. It might be a config issue?

1 hour ago, Ruedii said:

OK, after a heated conversation with the KIS devs, due to some issues with understanding whether the issue was a bug or a needed feature.   (I've found not to question matters of pride and call something a feature request when devs claim it's not a bug.  Makes them feel all warm and fuzzy being needed, instead of feeling like having made some sort of mistake.  Some dev's can't grasp the definition of bug-upstream workaround needed as not being their fault.) It seems the workaround for the ground issue is to adjust the mount node to have a little more offset from the collide.

For the plugs, you just need to cutoff their colider to not include the prongs.  (Simple enough).  However, I would recommend instead simply having the prong mounts connect with the pipes in question directly instead. This is probably a better option, although it requires more changes.

I would rather not get into anything with another dev, so I think the offset colliders is the best course of action. Since you've already gotten this far, would you be able to experiment and find the correct offset? Either of the newer experiments (gravimeter or magnetometer) would do, as their attach nodes are aligned exactly with the flat bottom of the collider.

I'm not sure what you mean by the second part of your bottom paragraph, but the collider is simply a box that is flush with the mating surface of the plug. The prongs are not represented in the physics at all.

[blu3wolf]

I cant recall if its been reported or not, but Im still getting explosions out of the gravimeter and the surface magnetometer. This was on Kerbin, next to the launch pad.

[CobaltWolf]

4 hours ago, blu3wolf said:

I cant recall if its been reported or not, but Im still getting explosions out of the gravimeter and the surface magnetometer. This was on Kerbin, next to the launch pad.

that's probably connected to what Ruedii mentioned, their colliders are flush with the attach nodes.

[vardicd]

4 hours ago, blu3wolf said:

I cant recall if its been reported or not, but Im still getting explosions out of the gravimeter and the surface magnetometer. This was on Kerbin, next to the launch pad.

 

16 minutes ago, CobaltWolf said:

that's probably connected to what Ruedii mentioned, their colliders are flush with the attach nodes.

Strangely enough I've found that if you attach something to the ground that doesn't explode, then attach an object that does explode to the top of that object, it no longer explodes. {not 100% fool-proof, not responsible for any actual parts exploding from this procedure. Warranty void if you've read this.}

[blu3wolf]

5 minutes ago, vardicd said:

 

Strangely enough I've found that if you attach something to the ground that doesn't explode, then attach an object that does explode to the top of that object, it no longer explodes. {not 100% fool-proof, not responsible for any actual parts exploding from this procedure. Warranty void if you've read this.}

Yeah, hence the attaching things to ground pylons/concrete bases. Used that to get the surface magnetometer report - although it still was not suitable for the contract I was trying to complete. Probably an issue with the contract pack in question (Kerbal Academy).

[Ruedii]

23 hours ago, CobaltWolf said:

Huh. They're set up the same in Unity. It might be a config issue?

I would rather not get into anything with another dev, so I think the offset colliders is the best course of action. Since you've already gotten this far, would you be able to experiment and find the correct offset? Either of the newer experiments (gravimeter or magnetometer) would do, as their attach nodes are aligned exactly with the flat bottom of the collider.

I'm not sure what you mean by the second part of your bottom paragraph, but the collider is simply a box that is flush with the mating surface of the plug. The prongs are not represented in the physics at all.

Yeah, the issue is that the nodes are flat to the bottom of the collider.  Either move the node down, or the collider up.  (Preferable would be to move the collider back from the visible surface enough to have a small lip of the object extend underground.)

As of some of the objects with more complicated colliders, simplifying them might be useful (combining the legs into a single solid piece as far as physics are concerned.  Nothing is small enough to go between them anyway.)

I've done quite a bit in config file wizardry before, but tweaking nodes is still something I've never done.  If nobody else is up to it I could give it a shot in a few days, but otherwise I'd greatly appreciate someone else do it that is experienced with it.

 

Edited August 3, 2016 by Ruedii

[CobaltWolf]

11 minutes ago, Ruedii said:

Yeah, the issue is that the nodes are flat to the bottom of the collider.  Either move the node down, or the collider up.  (Preferable would be to move the collider back from the visible surface enough to have a small lip of the object extend underground.)

As of some of the objects with more complicated colliders, simplifying them might be useful (combining the legs into a single solid piece as far as physics are concerned.  Nothing is small enough to go between them anyway.)

I've done quite a bit in config file wizardry before, but tweaking nodes is still something I've never done.  If nobody else is up to it I could give it a shot in a few days, but otherwise I'd greatly appreciate someone else do it that is experienced with it.

 

you assume the colliders are more complicated than they are. For most parts they are a cube, whose back face is flush with the plug face, the bottom is flush with the lowermost geometry, and the rest stretched to roughly match the boundaries of the geometry.

editing the nodes would be easy. In the node_attach line of the config, the second value in the sequence is the Y offset. Offset down from where they are - again, any of the newer parts will do as I know for sure their colliders are flush with the node. With some experimenting we could find out what the optimum offset is, and then use that to offset the colliders while keeping the current nodes.

[TwoThe]

So just that I understand this correctly: you fly this to a planet, place it on the ground, run the science experiment and sent it back home... and then the whole installation becomes practically useless, because you cannot repeat the experiments nor use it for anything else?

[cyberpunkdreams]

3 minutes ago, TwoThe said:

So just that I understand this correctly: you fly this to a planet, place it on the ground, run the science experiment and sent it back home... and then the whole installation becomes practically useless, because you cannot repeat the experiments nor use it for anything else?

Yes, just like the real Apollo SEP gear. The whole set-up is worth a lot of science though.

[blu3wolf]

3 hours ago, TwoThe said:

So just that I understand this correctly: you fly this to a planet, place it on the ground, run the science experiment and sent it back home... and then the whole installation becomes practically useless, because you cannot repeat the experiments nor use it for anything else?

Very nearly. Its worth pointing out that if you do leave it there, and you use RemoteTech, its an additional node in your comms network - and those are generally useful things to have, for redundancy.

[DStaal]

4 hours ago, TwoThe said:

So just that I understand this correctly: you fly this to a planet, place it on the ground, run the science experiment and sent it back home... and then the whole installation becomes practically useless, because you cannot repeat the experiments nor use it for anything else?

You can pick them up and re-run them in a different biome, of course.

[CobaltWolf]

4 hours ago, TwoThe said:

So just that I understand this correctly: you fly this to a planet, place it on the ground, run the science experiment and sent it back home... and then the whole installation becomes practically useless, because you cannot repeat the experiments nor use it for anything else?

not for long, friendo

 

 

[Ruedii]

10 hours ago, CobaltWolf said:

you assume the colliders are more complicated than they are. For most parts they are a cube, whose back face is flush with the plug face, the bottom is flush with the lowermost geometry, and the rest stretched to roughly match the boundaries of the geometry.

editing the nodes would be easy. In the node_attach line of the config, the second value in the sequence is the Y offset. Offset down from where they are - again, any of the newer parts will do as I know for sure their colliders are flush with the node. With some experimenting we could find out what the optimum offset is, and then use that to offset the colliders while keeping the current nodes.

So they already are greatly simplified.  That's good. 

You may want to push back the collides a tad further then from the visible bottom.  This is what KIS did with it's base anchor and it seems to work well.  Honestly, looking at the size of the parts on KSP, you should be able to push the collider up between 0.1 and 0.2m without a problem.

I would have to get Unity working and load up the parts myself to edit them.

OH, BTW, the node needs to be moved to the collider on the cardboard box.  I know it doesn't attach to anything, but the default height for drop is based on the bottom node, even if the bottom node is marked as unused.  It might be an idea to make the cardboard box radial attachable, but have very weak node attachment, making the node rip off on the slightest bump or shock.

[TwoThe]

9 hours ago, CobaltWolf said:

not for long, friendo

Sounds promising. And thanks for the answers.

I would find it better if this would work more like a fixed mobile science lab that generates science over time, because that is what this was used for and it would feel better as a reward then: "You assembled it, pushed the on button, and now you can dismantle everything again". And fixed in terms of that you cannot add more science to it later, so it runs out and becomes mostly useless eventually (except for the "report science from planet" type of contract maybe).

[CobaltWolf]

7 hours ago, Ruedii said:

So they already are greatly simplified.  That's good. 

You may want to push back the collides a tad further then from the visible bottom.  This is what KIS did with it's base anchor and it seems to work well.  Honestly, looking at the size of the parts on KSP, you should be able to push the collider up between 0.1 and 0.2m without a problem.

I would have to get Unity working and load up the parts myself to edit them.

OH, BTW, the node needs to be moved to the collider on the cardboard box.  I know it doesn't attach to anything, but the default height for drop is based on the bottom node, even if the bottom node is marked as unused.  It might be an idea to make the cardboard box radial attachable, but have very weak node attachment, making the node rip off on the slightest bump or shock.

I was saying that, I all that matters is the offset, that can be figured out by just editing the attach in the cfg to find out what works to further reduce the explosions. All that matters is finding that number, then when I get a chance I can edit the colliders and nodes as needed. You wouldn't have to install unity, just figure out the number by playing with the cfgs.

1 hour ago, TwoThe said:

Sounds promising. And thanks for the answers.

I would find it better if this would work more like a fixed mobile science lab that generates science over time, because that is what this was used for and it would feel better as a reward then: "You assembled it, pushed the on button, and now you can dismantle everything again". And fixed in terms of that you cannot add more science to it later, so it runs out and becomes mostly useless eventually (except for the "report science from planet" type of contract maybe).

The upcoming system is a fair bit more involved than that, and is still getting fleshed out. I'm not the one taking point on it, so I'd rather not blab too much and risk conveying false information. Once its at a point that its ready enough to show/discuss, they're free to post more information.

[cyberpunkdreams]

3 hours ago, TwoThe said:

Sounds promising. And thanks for the answers.

I would find it better if this would work more like a fixed mobile science lab that generates science over time, because that is what this was used for and it would feel better as a reward then: "You assembled it, pushed the on button, and now you can dismantle everything again". And fixed in terms of that you cannot add more science to it later, so it runs out and becomes mostly useless eventually (except for the "report science from planet" type of contract maybe).

Future developments notwithstanding, a big part of the point of it is to give your Kerbals something to do once they land somewhere, instead of just plant a flag, take a surface sample and bugger off home again. It takes an engineer to build and a scientist to use, which also makes planning missions more interesting. Agreed that it would be great if it accumulated science over time.

[CobaltWolf]

Those interested in the future of SEP should probably tune into my stream later tonight... (~2 hrs from this post)

[DiceCaller]

I don't know how well it works, because the Kraken ate my base.  I did have a scientist collect all the data and take into the lad at the base.  This was after I had transmitted the science back home. 

It is fun and nice to have something 'to do' besides just planting a flag.  Keep up the great work!

[SyberSmoke]

I have been wondering.  Any plans to add a longer term science experiment akin to Station Sciences experiments?

I ask because we go to a place, we set up the stuff.  We use the stuff...then...that is it.  And really with all this monitoring gear set in place...it strikes me as a opportunity to have short term experiments granting a low science.  Then have a long term experiment that takes a day or two, is worth a lot of science, and is designed to keep the kerbals at the site for a while.

A second thought could be that every so often, to a maximum, the control station could radio back science experiments connected to it.  So that setting them up and leaving them there could generate say a couple science per week up to a maximum giving the player a passive science income.

Thoughts?

[DStaal]

53 minutes ago, SyberSmoke said:

I have been wondering.  Any plans to add a longer term science experiment akin to Station Sciences experiments?

I ask because we go to a place, we set up the stuff.  We use the stuff...then...that is it.  And really with all this monitoring gear set in place...it strikes me as a opportunity to have short term experiments granting a low science.  Then have a long term experiment that takes a day or two, is worth a lot of science, and is designed to keep the kerbals at the site for a while.

A second thought could be that every so often, to a maximum, the control station could radio back science experiments connected to it.  So that setting them up and leaving them there could generate say a couple science per week up to a maximum giving the player a passive science income.

Thoughts?

I think they might be working on something like that.  And one of these days I'll get around to writing up a couple of ideas I have for MOLE that would also do something like that.

[blu3wolf]

4 hours ago, SyberSmoke said:

I have been wondering.  Any plans to add a longer term science experiment akin to Station Sciences experiments?

Thoughts?

My thoughts.... I think it might be worth reading the backscroll XD

[Javster]

This mod is great, I did a small test on the launchpad, and it worked well!

I forgot to refill the inventory of my Mun lander, so I couldn't try all the experiments on the Mun...

[Stone Blue]

23 hours ago, CobaltWolf said:

I was saying that, I all that matters is the offset, that can be figured out by just editing the attach in the cfg to find out what works to further reduce the explosions. All that matters is finding that number, <snip>. You wouldn't have to install unity, just figure out the number by playing with the cfgs.

@Ruedii Felbourn's NodeHelper .dll is a great tool for accomplishing this, quickly and easily...

Edited August 5, 2016 by Stone Blue