Kerbal Space Program Update 1.5 Grand Discussion thread

[netbumbler]

10 minutes ago, rudi1291 said:

Not sure if it was mentioned before... but whats with the lighting in the menu scene? It looks like there are two light sources

That would be Kerbol and Jeb's camera flash.... 

[juanml82]

1 hour ago, Shadowmage said:

That particular behavior is actually determined by Unity, and entirely out of SQUADs control.  More specifically, Unity does not allow for specifying local varying gravity as would be needed by KSP (e.g. two probes within the same physics bubble will have slightly different gravity directions), so KSP has to apply gravity forces 'manually'. AFAIK, if a rigidbody (such as used in every Part in KSP) is being subject to an external force, the physics simulation will not let them 'rest'.  Normally if a rigidbody is being effected only by gravity, but not moving (i.e. was landed), it will be put into a 'rest' state and physics calculations suspended until there is new external interaction (be it from collision, engaging an engine, whatever).

TLDR:  Unity physics simulation is not robust enough to handle KSP's physics requirements in an optimal fashion.

Just thought I would clear up a few things regarding Unity and physics...

Thanks for the explanation, but KSP specifically doesn't do the bolded part. No vessels within the physics bubble are ever in a 'rest' state.

[Deddly]

6 hours ago, AlamoVampire said:

@DeddlyOh and silly question the dlc is on Steam? 

Yes it's there

[Shadowmage]

37 minutes ago, juanml82 said:

Thanks for the explanation, but KSP specifically doesn't do the bolded part.

Because it can't.  It is not under the control of KSP -- it is part of Unity and PhysX.

Edit:  The reason it can't is because Unity does not support the per-object variable gravity setup needed.  Hence, not a problem with KSP per-se, but a limitation in Unity.  In this case, KSP is doing the best it can, given the engine it is using. (and no, I don't think any other commercial engines include variable per-object gravity vectors)

Edit2:  I'm not meaning to dismiss your concern, because it is perfectly valid -- the physics engine should do precisely such an optimization.

Edited October 18, 2018 by Shadowmage

[SkyRender]

Seems interesting, but I cannot for the life of me figure out how you create new vessels in Making History.  It just gives me a blank drop-down list when I try to do a creator built ship and there's no option anywhere to build a ship.

EDIT: Nevermind, figured it out.  I didn't expect the option to be at the bottom of the list...

Edited October 18, 2018 by SkyRender

[Johnny Wishbone]

Aerodynamics definitely seem different with this release. I have a fairly simple aircraft that used to glide like a champ in previous releases. Now, it drops like a brick. If I had to guess, it seems like it is generating more drag, causing the glide slope to be much more severe than it used to be.

[cfds]

2 hours ago, Shadowmage said:

In this case, KSP is doing the best it can, given the engine it is using.

Except just merging the parts into a single rigid body when the situation calls for it...

[Shadowmage]

5 minutes ago, cfds said:

Except just merging the parts into a single rigid body when the situation calls for it...

First they would have to remove the mandate that parts be joined together by flexible joints.  Likely they would have to come up with a replacement system to allow for aero/physics based RUD that didn't rely on the joints breaking.  They would also have to find a way to specify the proper mass distribution / inertia tensors for the single rigidbody that accurately reflected its 'bunch of separate parts' configuration.

Certainly just that one change would eliminate a large portion of the physics load, instead of simulating one rigidbody per-part, you would have one per-vessel.  Imagine being able to run a 1000+ part vessel with no noticeable slowdown?  How about 5 of them at once?  I've often been tempted to try and do such a thing through a mod -- but it would be a major undertaking, and I don't think that access is available for mods to adjust all the systems that would need to be changed.

On a more positive note, I would like to give kudos to whomever at @SQUAD did the texturing work for the new Stayputnik -- when you plug the existing textures into a fully functional shader (and don't use dx9), the results are impressive.  Truly well done   It looks like the existing textures already include a smoothness map in the spec.a channel, and plugging it all into the Unity Standard (Specular) shader, you get a level of texturing artwork unseen anywhere else in KSP.  There is a ton of detail in that unused smoothness map -- scratches, smudges, dirt/dust; someone spent some time and effort to create the map, but it just sits around unused by the stock shader.

[Deddly]

35 minutes ago, Johnny Wishbone said:

Aerodynamics definitely seem different with this release. I have a fairly simple aircraft that used to glide like a champ in previous releases. Now, it drops like a brick. If I had to guess, it seems like it is generating more drag, causing the glide slope to be much more severe than it used to be.

This is partly correct. I believe it's confirmed that there is a bug in lifting bodies.

[Geonovast]

3 minutes ago, Shadowmage said:

I didn't realize the Death Star came with a display stand.

[KevinJohn3D]

7 minutes ago, Geonovast said:

I didn't realize the Death Star came with a display stand.

That's because that's the finished second Death Star. The rebels destroyed it however before the empire could finish construction. It's all in the novels...

[Catatau_27]

Good afternoon everyone, I had good impressions of version 1.5.0, it seemed much easier to take off or land with any plane, I have not done tests in space, but even for me I am by far the worst pilot in this whole forum , I was able to land several times without losing control, blowing up or losing parts of my planes. The mod: adjustable landing gear, is working perfectly in this version, without entering the floor of the track, I realized that the jet engines do not need much to work, as it was in the previous version where it was necessary to have more track and more force to leave the floor.

[TeslaPenguin1]

2 hours ago, Johnny Wishbone said:

Aerodynamics definitely seem different with this release. I have a fairly simple aircraft that used to glide like a champ in previous releases. Now, it drops like a brick. If I had to guess, it seems like it is generating more drag, causing the glide slope to be much more severe than it used to be.

All parts that generate body lift have the lift vector rotated by 90 degrees, so it just makes more drag. Wings are fine, but body lift (like from mk2)? Nope.

[Johnny Wishbone]

2 hours ago, Deddly said:

This is partly correct. I believe it's confirmed that there is a bug in lifting bodies.

 

13 minutes ago, TeslaPenguin1 said:

All parts that generate body lift have the lift vector rotated by 90 degrees, so it just makes more drag. Wings are fine, but body lift (like from mk2)? Nope.

Except that my craft doesnt use lifting bodies. Its a MK1 cockpit with structural fuselages, FL-200 fuel tanks, a terrier engine, and some wings. Like I said, it used to glide really well in previous releases. Now, not so much.

Edited October 18, 2018 by Johnny Wishbone
Spelling

[Poodmund]

With regards to the above.

[Johnny Wishbone]

2 minutes ago, Poodmund said:

With regards to the above.

Okie dokie.

Unfortunately, I wont have a chance to patch and test until tomorrow, but I will follow up in this thread if this hotfix addresses my issue

Thanks

[AngryKitty]

1.5.1 definitely fixes the aero issues...and the secret OS X patch a couple days ago definitely fixed up performance. (Actually not sure what happened there, but CKAN told me I went from 1.5.0 to 1.5.0.2something...) 

Looks great so far!

[Deddly]

1 hour ago, Johnny Wishbone said:

 

Except that my craft doesnt use lifting bodies. Its a MK1 cockpit with structural fuselages, FL-200 fuel tanks, a terrier engine, and some wings. Like I said, it used to glide really well in previous releases. Now, not so much.

The parts you mention also work as lifting bodies, although not as much as MkII parts. Press F11 to see them in action.

[Johnny Wishbone]

39 minutes ago, Deddly said:

The parts you mention also work as lifting bodies, although not as much as MkII parts. Press F11 to see them in action.

Thank you. I’m well aware of how to turn on the overlays. :)

However, I was not aware that basic fuel tanks and round structural fuselages now generate lift. The MK2 tank parts clearly state the lift they produce whereas the regular round ones (ie: FL-T200) dont show any information about producing lift. 

[JPLRepo]

1 hour ago, AngryKitty said:

1.5.1 definitely fixes the aero issues...and the secret OS X patch a couple days ago definitely fixed up performance. (Actually not sure what happened there, but CKAN told me I went from 1.5.0 to 1.5.0.2something...) 

Looks great so far!

there was no secret OSX patch.

[MikeO89]

For me without a doubt the most noticeable difference between 1.5 and 1.4.3 is the performance hit. On my system with same craft, same mods, same saves, the performance hit is pretty significant. Load times for me in 1.5 feels close to double what 1.4.3 was. Map takes longer to access. No matter what might be good about 1.5 (and I really haven't found anything to note yet), if the performance takes a noticeable hit, it kind of wipes that out. Squad has had quite some time now to work on performance. I think that's where the focus should be.

Edited October 19, 2018 by MikeO89

[Lisias]

3 hours ago, MikeO89 said:

On my system with same craft, same mods, same saves, the performance hit is pretty significant. Load times for me in 1.5 feels close to double what 1.4.3 was. Map takes longer to access. 

You should be making comparisons with unmodded KSPs first.

Using modded installs, you are comparing mods, not just KSP. and since such mods are unlikely to be all "certified" for 1.5, the performance hit can be due exceptions being thrown regularly. It essentially what happens when I mess up on mods I change.

Edited October 19, 2018 by Lisias
Krakens count the typos!

[MikeO89]

Point taken. I don't play unmodded KSP though so no reason for me to do that comparison cause it wouldn't mean anything to me. I don't even have an unmodded version of KSP, shows what a mod wacko I am. I play what I compared. For KSP purists my comparison might be pretty worthless. Not everyone is a purist though. I'm so un-clean.

Edited October 19, 2018 by MikeO89

[Deddly]

2 hours ago, MikeO89 said:

Point taken. I don't play unmodded KSP though so no reason for me to do that comparison cause it wouldn't mean anything to me. I don't even have an unmodded version of KSP, shows what a mod wacko I am. I play what I compared. For KSP purists my comparison might be pretty worthless. Not everyone is a purist though. I'm so un-clean.

I think what he means is that you should compare the performance between the two stock versions. That way, you will know for sure whether or not 1.5 really is at fault for your performance loss. 

You don't have to play in stock apart from testing the performance, but it's the first step in finding what is causing the problem. 

[NoobTool]

I've noticed a pretty substantial performance hit in a clean stock install. What's more, the performance changes seem related to how many/what type of engines are activated (NERV's seem to be the biggest offender). This, along with the fact that burn-time is being calculated incorrectly for NERV-powered craft, leads me to believe that the issue is related to the new extended burn-time calculator.