Anth

Could Squad have fixed the problem? You know a lot more about this from looking at that link. Also the main memory leak was directly proportional to the size of the game save. Was it all the parts of the entire save even if it wasn't in scene? My bug report https://bugs.kerbalspaceprogram.com/issues/25218 for game save size testing for the most part used small crafts in scene and had huge crafts that were never loaded and the memory leak still happened. At one point KSP1 would have obvious stuttering problems when the garbage collection was happening regardless of how much memory I had available. And then the stuttering would get more and more frequent. I don't know which version that stopped happening and then after that it was just obvious with the increase in memory and the longer and longer load times.

That particular memory leak is caused because KSP loads/saves at most scene changes and the memory leak is directly proportional to the size of the save file. I have the following bug report https://bugs.kerbalspaceprogram.com/issues/25218 where I did 55 different actions with graphs of recorded memory I manually took from the windows task manager using different sized saves over different versions of KSP (By the way it started when Squad added the so called helpful garbage collection that Unity does in 1.2.0) Examples of where I concluded from my bug report this was occuring: First time its loaded from the main menu Everytime going from a loaded scene to the KSC screen Everytime going from the KSC screen to the SPH/VAB Everytime that a craft is launched from the SPH/VAB Everytime that a craft is loaded from the tracking station Everytime going from a loaded scene to the tracking station Everytime going to the KSC Screen I then tested this theory with Subnautica which is also a Unity game. A Subnautica save normally is loaded once by a player when they first start up the game but I loaded a save, then went back to the main menu and loaded it again and back to the main menu etc and there was evidence that it was also suffering from the same problem but its not loaded enough times to be noticed. Squad said this was not fixable though 1.1.3 wasn't suffering from the problem. Its a major thing I will be looking for in KSP2. I doubt that Intercept Games will allow this to happen. They will figure out a work around like we do when navigating KSP1 bugs to get what we want out of the game. Also I am wanting to see if the move from the SPH/VAB to scene is instantaneous in KSP2 because the only times the game should need to be loaded is for revert to launch and when loading a quicksave and loading the save for the first time. The KSP1 debug menu has 'set orbit' and 'set position' which can move crafts within the Kerbol system without loading or saving. Oh @K^2 and @Gotmachine posting while I was creating my response. Hey guys

I will post my results for sure. A thread like you are suggesting is a good idea. I wonder if Intercept Games could have a benchmark built into KSP2 that might make that easier.

60fps would be nice

I am not sure everyone has computers that are at that level. But whenever the required specs for KSP2 come out they will hopefully tell us in a weird way where the people's computers are in general for gaming. Preferable yes. If KSP2 ended up only being able to play on the most recent computers even if they are mid tier I think they would be losing customers. I am hoping it will play at least ok on my i7 4770K and 1080GTX. On my i9 9900K and 3070ti I expect it to run amazingly well once they have optimized it properly. The KSP2 Artemis video showed some performance issues so I don't know exactly what to expect on day 1 of the release.

I stated previously in this post that I would be testing it on: Intel i7 4770K (4 cores) (I want to see how this performs with the onboard graphics only as well) Nvidia 1080GTX 32GB of RAM My main computer is the following: Intel i9 9900K (8cores) (I want to see how this performs with the onboard graphics only as well) Nvidia 3070ti 32GB of RAM It will be interesting to see how they compare. The i9 9900K might be able to compensate for any performance issues but hopefully the i7 4770K will show issues in a very obvious way. Wait. the Ryzen 7 3700X isn't that much less in performance than my i9 9900K. There is no way that the majority of people have the equivalent of a Ryzen 7 3700X level CPU. KSP2 is going to have to be able to perform on a lot less than that.

Its going to be fun comparing KSP1 to KSP2 when we get it into our grubby hands. One thing I will have open on my other screen will be the task manager showing me KSP CPU usage and memory usage. On the memory side I will be watching to see if scene changes (caused by the save file being loaded and/or saved at most scene changes) are still contributing to a severe memory leak after Squad said that it was a Unity issue and couldn't be fixed.

Nate Simpson talks about parallelization in the PC Gamer interview but nothing specific enough about parts physics. Should be easy enough to test on my old computer with a high part count test. Maybe they have a way of doing it on more than one core? Or one core is enough for complete dedication to the parts physics and then focus the other cores on other tasks?

KSP2 wont be a single core game though. I guess we wont see how much it uses the CPU until early release though.

Well hopefully its good enough to show me potential issues with the game performance. Running KSP2 on my better computer might hide issues with the game. 4ghz and 8 cores? that should be fine for KSP2 right? What GPU do you have?

How old is too old for a computer that can run KSP2 adequately? I have an older computer that I haven't used in a while that I want to test KSP2 on which is the following: Intel i7 4770K Nvidia 1080GTX 32GB of RAM is that too old?

I have made many screenshots in KSP1 but photomode like we see in other games would be great. Also I like the idea of seeing a Kerbal doing a thumbs up from within a capsule and different emotional expressions that are controllable. (I am sure others have mentioned this on the overall forums, I just cant find it in the suggestions.

I have never had the problem you are talking about. As for stopping its fast roll in space turn on timewarp as long as its not under some sort of power then the craft will stop spinning.

You are right it does take course corrections to get just the right positioning for a Kerbin atmospheric trajectory from the Eve SOI and sometimes I have to make adjustments in solar orbit.

You haven't ever done an Eve SOI to Kerbin atmospheric entry?

Seattle is a cold city? I imagine soup is something consumed when its cold.

From close perhaps, but imagine half way between the Mun and Kerbin orbit. Or having a maneuver node that is multiple orbits away.

I hope that something like the trajectories mod will in KSP2 as stock but also I wouldn't mind knowing if a craft is going to land during the day or night. Probably hard to predict over years if the impact on Tylo was going to be in 4 years and the craft is still within Kerbin SOI but if the craft was closer?

Sweet thank you.

@Ghostii_Space Good Dev Diary to see. I look forward to the early access process and the back and forth between us the community and yourself and Intercept Games. It will be fun. I have two questions Will we have a bug tracker for the early access process? I did quite a few bug reports but also interacted with others bug reports to figure out (confirm) problems. It was a fun experience. The feedback system that is talked about being connected to the launcher, do we the community get to interact with others feedback like a public forum or will the feedback be one way?

We will find out soon enough if what we are saying has any major affect on features that seem to be completely fleshed out. I hope it ends up feeling a bit like a community effort than just giving blind feedback to Intercept Games.

If our expectations are lower, but then KSP2 performs better when we get our grubby hands on it then that will be better than the other way around right?

Yeah that would a concern. What I would really like to do is progress far enough in my KSP2 save that its ready for the next steps in the roadmap rather than have to start from scratch in my save. It isn't like everything after early access are going to be built from the ground up though. They appear to have been working on all features quite a bit. So we might be lucky that crafts will stay compatible with all updates.

Thorough testing is what should have happened for the maneuver mode in the first place. For them to get the phase angle blatantly wrong means they didn't test it at all. I tested it with Kerbal Engineer. But perhaps you are right. Its a little worrying that code would not be able to be fixed with a +1 though. Maybe.

@Lisias So looking at everything you have said basically Take Two had no choice but to do a complete reset giving KSP2 to another company. Then they possibly looked at Squad's way of doing things to see if their employees would be capable under new management. Which they are or they never would have ended up working on KSP2. Its good that they are working on KSP2 so they can give feedback to keep KSP2 feeling KSP like. Also they might know issues we have no idea about with KSP1 so as to counter the same happening in KSP2. Well it was lucky you kept your job. I have seen people fired for doing what you did even if the proposal of changes would have increased production.