Jump to content

[1.12.1] JNSQ [0.10.0] [23 Sept 2021]


Galileo

Recommended Posts

JNSQ Patch

Many of you have undoubtedly experienced a terrain bug while playing JNSQ or other planet packs that looks like this:

brgPghj.jpg

Thanks to @R-T-B, the cause of this has been identified.  I've written a patch for JNSQ that should fix it.  Just download the following zip file, select the version of KSP that you are currently using, and install the patch as indicated.

Download JNSQ Patch

Note that this patch is different from the one provided by R-T-B in the preceding post.  R-T-B's patch reverts several celestial bodies to using the pre-1.8 terrain shader.  My patch changes the templates of those bodies so that the new terrain shader can remain enabled.

WARNING...

There is evidence suggesting that this patch can adversely affect contracts.  For instance, it has been reported that Mun contracts prior to the patch were converted to Moho contracts after installing the v1.8.1 patch.  Unfortunately there is no way to change the patch to prevent this.  Therefore, it is recommended that you complete, or cancel if you can afford to, all accepted contracts prior to installing the patch.  New contracts generated post-patch appear to be OK.

 

Edited by OhioBob
Link to comment
Share on other sites

There could be some unintended consequences with changing templates, but hopefully nothing major.  One thing I know that happened is when I switched templates I just removed all PSQmods from the new template.  It was just much easier and less time consuming to do it that way.  One consequence of that is it removes the Randolith from spawning on those bodies.  Not that big of a deal but it is something if you like to hunt anomalies.  Hopefully there will come a time with future KSP releases that a patch will no longer be needed.

Link to comment
Share on other sites

On 10/2/2020 at 7:45 AM, alberro+ said:

Switching today back to JNSQ from KSRSS. I literally can't live without this.

oh man, your recent posts reminded me to try to get KSRSS working again ... but no one seems to know how to fix my rescale issue. trying to get these things working can be ... "challenging" to say the least.

On 10/1/2020 at 4:47 PM, OhioBob said:

Many of you have undoubtedly experienced...

For as much as I saw this bug in lots of planet packs, I hardly ever saw it in JNSQ, which is where I spend like 80% of my KSP time. No idea why! But thank you for the patch, this is awesome!

Link to comment
Share on other sites

On 10/1/2020 at 4:47 PM, OhioBob said:

JNSQ Patch

Many of you have undoubtedly experienced a terrain bug while playing JNSQ or other planet packs that looks like this:

brgPghj.jpg

Thanks to @R-T-B, the cause of this has been identified.  I've written a patch for JNSQ that should fix it.  Just download the following zip file, select the version of KSP that you are currently using, and install the patch as indicated.

Download JNSQ Patch

Note that this patch is different from the one provided by R-T-B in the preceding post.  R-T-B's patch reverts several celestial bodies to using the pre-1.8 terrain shader.  My patch changes the templates of those bodies so that the new terrain shader can remain enabled.
 

Ok, waitaminute - this is a patch to fix the... "patch"? :-D 
That's very meta. (and appreciated)

Link to comment
Share on other sites

2 hours ago, GameTourist said:

I just noticed that this mod has two moons named Amos and Prax... and I just love this mod that much more :wub:

Sorry but I don't know what you are referring to.  If the Amos and Prax names are used elsewhere, then it's just a coincidence.

Link to comment
Share on other sites

On 10/1/2020 at 9:16 PM, OhioBob said:

There could be some unintended consequences with changing templates, but hopefully nothing major.  One thing I know that happened is when I switched templates I just removed all PSQmods from the new template.  It was just much easier and less time consuming to do it that way.  One consequence of that is it removes the Randolith from spawning on those bodies.  Not that big of a deal but it is something if you like to hunt anomalies.  Hopefully there will come a time with future KSP releases that a patch will no longer be needed.

I'm playing 1.8.1 on a Mac.

Installed this patch, as the patchwork quilt look bugged me a bit.  ( It's only aesthetics, but that's part of the JNSQ appeal beyond the re-scale )

I'm seeing a *massive* impact on performance though.  Game loads, but is massively slow to the point of unusability.  Entering the Tracking Station the planets are very dark in apperance, and any attempt at zoom in or out, results in the dreaded 'spinning beach ball'.  Nothing jumps out in the log, just the usual scene transition messages I'm used to seeing.

It appears to be due to memory, as the process rapidly consumes all memory until it starts thrashing.  I have 16G of RAM, which has been sufficient to play KSP ( I quit everything else, so KSP and a terminal window are the only things running ).

Does this patch potentially drive different memory behavior of caching textures that would result in this increase?  Or drive the use of different shaders?

Removing the patch restores playability for me.

Link to comment
Share on other sites

@coredumpster, I don't know why anything in the patch would cause the problems you're describing.  Unfortunately I can't test it because I don't have a Mac.  Sounds like you're just better off not using the patch, or consider upgrading to 1.9.1.

(edit)

I just tested it on my PC in 1.8.1 using the patch and everything seemed to be working fine.
 

Edited by OhioBob
Link to comment
Share on other sites

11 minutes ago, OhioBob said:

@coredumpster, I don't know why anything in the patch would cause the problems you're describing.  Unfortunately I can't test it because I don't have a Mac.  Sounds like you're just better off not using the patch, or consider upgrading to 1.9.1.

Yeah, it's not a huge ordeal, as I've played without it for this long, and everything else is awesome.  :)

I was holding off upgrading until 1.10.x stabilizes a bit more, and I'm almost done with my end-game play-out on 1.8.1 ( first end-game playthrough on JSNQ ).

I just need to crack Eve.  I've got a vessel that can make it back to orbit from near sea level.  My problem is keeping the vessel from burning up on entry!  Eve atmo and orbital speeds are tough in JNSQ.

 

Link to comment
Share on other sites

21 hours ago, OhioBob said:

I just tested it on my PC in 1.8.1 using the patch and everything seemed to be working fine.

Can confirm that on 1.9.1, the performance with and without the patch is identical. Same is true for other planet packs where I applied similar patches to the JNSQ one. 

edit: i should mention i'm on windows

Edited by OrbitalManeuvers
Link to comment
Share on other sites

On 10/5/2020 at 1:08 PM, coredumpster said:

I'm playing 1.8.1 on a Mac.

Installed this patch, as the patchwork quilt look bugged me a bit.  ( It's only aesthetics, but that's part of the JNSQ appeal beyond the re-scale )

I'm seeing a *massive* impact on performance though.  Game loads, but is massively slow to the point of unusability.  Entering the Tracking Station the planets are very dark in apperance, and any attempt at zoom in or out, results in the dreaded 'spinning beach ball'.  Nothing jumps out in the log, just the usual scene transition messages I'm used to seeing.

It appears to be due to memory, as the process rapidly consumes all memory until it starts thrashing.  I have 16G of RAM, which has been sufficient to play KSP ( I quit everything else, so KSP and a terminal window are the only things running ).

Does this patch potentially drive different memory behavior of caching textures that would result in this increase?  Or drive the use of different shaders?

Removing the patch restores playability for me.

Hi,

I'm using a 1.8.1 version under Windows with JNSQ and have installed the patch with no issues. I also have 16GB memory and did have memory issues a few months ago when I installed the Scansat update. With JNSQ, you can run out of memory, even with 16GB. There was a very small increase in mem usage after my update and that was enough to make it not run anymore. I increased my swap to 12GB total and lowered my background memory usage and all seems well now. I found the memory issue by running the Windows Task Mgr/Resource Mgr in my second window.  It seems like even a small increase memory can get you over a threshold, so you may wish to look into that with the corresponding MAC tools.

Edited by Gilph
clarified I have Windows
Link to comment
Share on other sites

On 9/26/2020 at 12:50 PM, Akira_R said:

So I'm having an issue with performance around added launch sites that seems to be related to terrain scatters. I am running a pretty heavily modded install. When I launch a plane from any of the airports around kerbin that JNSQ adds my frame rate tanks to 5fps. I am launching via the KCT menu but I don't think that should have much impact on this. If I fly far enough away (roughly 40km or so) from the launch site the frame rate comes up to 10-11. If I then quit and reload I'm up in the 20s but it will rapidly drop to 10-15, not ideal but doable. Flying back causes the frame rate to drop back to single digits, interestingly if I fly low and approach the center of the airport area frame rate comes up to like 10-15 in a very small area around the center and then craters again once I leave. There is no NRE or other exception spam. My system is a OCed i7 8700k, 32GB RAM and a gtx1080, the GPU is idling, CPU is around 20% utilization and none of the individual threads show utilization above that.

I tried turning terrain scatter off (it was set to around 45%) and saw a massive increase in frame rate, 40fps while flying around and around a 36fps average when returning to the airport. Searching around I am seeing that some performance hit is expected with scatters on especially with kopernicus giving them collision. But this impact seems excessive? Especially since I'm on a pretty beefy system and in terms of resource utilization it is hardly even being taxed. Any one have any idea what is going on here?

log if it'll help: https://gist.github.com/AkiraR/dbf2cee8b8e4ab47a7394544903f4f2d

I've been investigating this issue with JNSQ.  It seems the collision detection it does on scatters is very very expensive.  The best thing you can do is turn down scatterer percentage.  I'd advise nothing higher than 25%.

I tried to improve the performance of this section of Kopernicus in the code.  My best result was fairly good in release 37 of the bleeding edge.  But I had to revert it because it hurt some other performance scenarios outside of JNSQ.  If you are only using JNSQ, you may still try it.  It may give you a nice uplift.  Some day I hope to bring those optimizations back.

Edited by R-T-B
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...