Kerbal Space Program 1.12.3 is live!

[Anth]

On 2/2/2022 at 4:12 AM, Starwaster said:

That's just a KSP thing resulting from math errors with the craft as it is continously serialized/deserialized over time.

The game was designed to save the crafts original part coordinates only. That is evident from the persistent/save file.

If the game was supposed to remember the in real time stress of an attachment point location of each part of a craft then the game would need to save the following:

• The amount of force that is being applied to any given part at the time the game is saved (which KSP1 doesnt)
• The zero point where the attachment point is at rest. (KSP1 actually does save this)
• The twisted position of the part at the time the game was saved (knowing that this is NOT the zero point)

KSP 1 knows/assumes only the following from the save file:

• All parts are NOT under any force at all
• All parts are at rest when loaded into a scene (then physics is applied afterwards)
• All parts are in their original positions as they were in the editor

What happens with drift issues:

• The stressed position is recorded into the persistent file or save file
• The zero point of where the attachment point is at rest is moved to the stressed position that is recorded in the persistent file or save file

Then on reloading/revisiting the same craft that happens again and again, causing the craft to warp further each time.

The main cause of robotic drift has nothing to do with math errors at all. Why do people keep saying it is?

Edited February 3, 2022 by Anth12

[Lisias]

7 hours ago, Anth12 said:

The main cause of robotic drift has nothing to do with math errors at all. Why do people keep saying it is?

By many reasons.

Some, because they don't know better and extrapolate known problems of the past into new problems found nowadays.

Others, because people they trust are saying this B.S. and so they repeat it too.

A few ones by pure incompetence and/or lack of skills. Or perhaps just people on someone's payroll looking for scapegoats to mask the mistakes.

I found at least ONE situation where this "deformation" is happening, and it involves changes at Flight Scene on a craft under stress where AutoStruts (to the Grand Parent for sure, others to be tested) is applied on the parts currently under stress. And this is happening since KSP 1.2.0, when the Auto Struts were first introduced : it's a flaw on the initial implementation that was never detected before (or perhaps it was?  ).

You may like to check my findings documented on this comment and forward on github .

Edited February 3, 2022 by Lisias
grammars. Don't you hate this thing?

[Serenity]

14 hours ago, Starwaster said:

It's not simulating reality.

It's broken. It's a bug. It's defective.

You can never tell with @kerbiloid comments, it is a mystery what he really means.

Serious, sarcastic, both or whatever else? I am always confused, in this case i am going to go with sarcastic/funny  

But yeah it is a bug and it is broken but i am thankful we got that workaround update on what it seems to be a VERY tough or impossible problem to solve 100%.

Edited February 3, 2022 by Serenity

[4D4850]

On 1/10/2022 at 12:34 PM, Lisias said:

IMHO, Private Division should spend a little more money with at least one good developer to tackle down the worst problems on the bug track

I get that the answer to this is probably no, but would it be possible for the community to create a patch for the game to fix major bugs, sorta like the Skyrim community patch?

[Lisias]

3 hours ago, 4D4850 said:

I get that the answer to this is probably no, but would it be possible for the community to create a patch for the game to fix major bugs, sorta like the Skyrim community patch?

Not without risking taking this kind of crap.

Even by being legal (what's currently under dispute - and I want to emphasise the company suing the modders), it theoretically violates Forum ToS and, so, you are prone to have the thread deleted (or even your account locked) at any time by rules violation. And I will ignore the necessary competence and skills needed to diagnosing problems correctly - something very important when you aim to fix problems without creating new ones.

Some problems are buried deep on the KSP guts, and it's unlikely that one could work around these bugs without access (and changes) on private methods (violating the Forum ToS) and without decompiling the binaries into source code (violating the Forum ToS and the game EULA - and, perhaps the Law).

As the situation stands, any effort following this line of research may be tolerated or even silently endorsed by the current Administrators and Managers, but people come and go on Companies and any change of personal can gave you headaches later - the kind of headache a professional on the field does not want to have. I surely don't.

What is left is the Clean Room Reverse Engineering (the approach I took on KSP-Recall), but instead of having someone decompiling code for me (what can be a source of problems) and writing specifications, I do tests and more tests (we call this Black Box Testing) on many APIs and manually crafted Craft Files and Savegames (Exploratory Tests) trying to infer such specifications - but this is time expensive. And having less than competent people injecting noise all the time on the public discussions about the problems hinders even more the efforts of the (very few) people really wanting to do any help here.

Edited February 3, 2022 by Lisias
bragging a bit my skills. :)

[4D4850]

Just now, Lisias said:

Not without risking taking this kind of crap.

Even by being legal (what's currently under dispute - and I want to emphasise the company suing the modders), it theoretically violates Forum ToS and, so, you are prone to have the thread deleted (or even your account locked) at any time by rules violation. And I will ignore the necessary competence and skills needed to diagnosing problems correctly - something very important when you aim to fix problems without creating new ones.

Some problems are buried deep on the KSP guts, and it's unlikely that one could work around these bugs without access (and changes) on private methods (violating the Forum ToS) and without decompiling the binaries into source code (violating the Forum ToS and the game EULA - and, perhaps the Law).

As the situation stands, any effort following this line of research may be tolerated or even silently endorsed by the current Administrators and Managers, but people come and go on Companies and any change of personal can gave you headaches later - the kind of headache a professional on the field does not want to have. I surely don't.

What is left is the Clean Room Reverse Engineering (the approach I took on KSP-Recall), but instead of having someone decompiling code for me (what can be a source of problems) and writing specifications, I do tests and more tests (we call this Black Box Testing) on many APIs and manually crafted Craft Files and Savegames trying to infer such specifications - but this is time expensive. And having less than competent people injecting noise all the time on the public discussions about the problems hinders even more the efforts of the (very few) people really wanting to do any help here.

Ok. That's unfortunate.

[R-T-B]

I'd just like to point out here that KSPCommunityFixes does in fact, exist.

It may exist despite the threat of potential legal banishment, but it does.  I'd encourage collaboration with them on stock issues.

Edited February 4, 2022 by R-T-B

[Andrewson]

10 June 2022  ksp 2 

[AVaughan]

57 minutes ago, Andrewson said:

10 June 2022  ksp 2 

Source? 

Edited February 5, 2022 by AVaughan

[Anth]

Its a joke. There's a joke going on the reddit KSP discord about it.

[Vanamonde]

Some posts removed. Please don't bring disputes from elsewhere on the internet to this forum. 

[BigStar Aerospace]

Great, when's 1.11.1 going to come to Enhanced Edition? Been a year or somewhere around there right?

Wait, wait, sorry, two years? Wow, sure flew by fast! Hope KSP 2 does better with that, because to be honest, it's getting really frustrating, I don't think I've seen even a bug update for more than a year, and KSPEE has allot of bugs.

[Anth]

I am pretty sure that KSP2 will be released on PC and console at the same time so hopefully that means that whatever updates afterwards will be consistently matched.

[Anth]

 

2 hours ago, BigStar Aerospace said:

Wait, wait, sorry, two years? Wow, sure flew by fast! Hope KSP 2 does better with that, because to be honest, it's getting really frustrating, I don't think I've seen even a bug update for more than a year, and KSPEE has allot of bugs.

Below is a quote from above forum post. @linuxgurugamerwas one of group of people that was invited to see Star Theory  to discuss the development of KSP2.

On 9/2/2019 at 1:54 AM, linuxgurugamer said:

Having spoken to the developers, their answer is this:

Star Theory is doing all the development on KSP-2.  An outside company is codeveloping the console versions at the same time is Star Theory is developing the pc version.  There are a few employees in Star Theory who are working on the console version.  They do recognize the difference between a computer and a console, which is why the console versions will not be released at the same time as the pc versions.  And they do play it on the console as well.

My take on this is that they recognize the mess that KSP on consoles is, and are doing their best to avoid the problems that happened.

Even though Star Theory is no longer part of the development, the major players like @Nate Simpson went to Intercept Games so I think the information will still be current.