Lisias
-
Posts
6,484 -
Joined
-
Last visited
Content Type
Profiles
Forums
Developer Articles
KSP2 Release Notes
Bug Reports
Status Replies posted by Lisias
-
Hi!
I have some news for you on Recall's thread:
Not the news you would like, I still don't know whats happening - I could only confirm this is not a problem being caused by Recall neither TweakScale.
I'm trying to figure out what to do next.
-
Just to finish the report, as I had diagnosed things later and my last message here left implied that KSPIE and/or IT were at fault somehow:
- It's a KSP's limitation : the IPartCostModifier uses float , and so are prone to loss of precision with really expensive (or cheap) parts.
-
KSPIE had no role in the problem, other than triggering it.
- IT, ditto. It only happens that these parts are expensive enough to get screwed by the float squashing problem.
- A new PartModule were added on KSP-Recall to handle the situation.
- All the other problems detected were just noise, they didn't affected the outcome.
-
-
Hi!
I have some news for you on Recall's thread:
Not the news you would like, I still don't know whats happening - I could only confirm this is not a problem being caused by Recall neither TweakScale.
I'm trying to figure out what to do next.
-
Hi!
I concluded the report:
There's something fishy on the IT part config. Since I don't know the KSPIE PartModules, I can't say what, but hey… Now we know where and how.
You will need help from the IT guys, and probably from the KSPIE's maintainer too in order to identify exactly why only IT parts are borking.
-
-
Hi!
I have some news for you on Recall's thread:
Not the news you would like, I still don't know whats happening - I could only confirm this is not a problem being caused by Recall neither TweakScale.
I'm trying to figure out what to do next.
-
Nope. TS is the trigger for the problem, as it changes the mass and cost of the part.
I was thinking perhaps a missing use case on KSP-Recall, but apparently it isn't neither.
Apparently it's something on the part config itself, I'm guessing some missing or improper value on something that, so, misbehaves on that part only and then induce KSP-Recall to bork - whatever is happening, is screwing with `Refunding` in a way that I still didn't understood.
My best guess, at this moment, is that whatever is getting screwed, is screwed while its code for the IPartCostModifier is being called and so when `Refunding` calls it to calculate the part costs it gets a bogus value.
Now my problem is to understand what and how (as when it appears to be obvious) and then see if there's something I can do on code, or if I will need to publish a patch for IT.
-
-
Hi!
You recently asked on the 1.12.3 release announcement about a real fix for the parts drifting:
On 1/12/2022 at 11:52 PM, Staticalliam7 said:Ah, makes sense. Is there a mod that corrects this, or is it too deep within the game's code to mess with (without losing your mind)
I think you will be interested on my findings on KSP-Recall:
A "fix" is still far away, but at least I found my way on the problem.
Cheers!
-
Found the problem on your MM patch!
TL;DR: shove a :FINAL on the patch and it will work.Don't have an answer for this yet, it's working hours. I will try to explain that at night!
-
About this post...
2 minutes ago, FreeThinker said:Make sure when reinstalling KSPIE, you remove Interstellar_Redist.dll from all Plugin folders (both WarpPlugin and PhtotonSailor) and it should only be present in the GameData folder
I also corrected the CKAN file for this change.
It was my understanding that KSP 1.12 changes on the DLL loading mechanism had solved this issue...
This is essentially what MM WatchDog prevents about MM, so what do you think to extend it a bit and check for multiple Interstellar_Redist.dll too and issue a warning if more than one is found? Ideally it should warn about its absence when it should be there too, but this is a bit more tricky (but not that much).
As it appears, this problem is going to stay...
-
About this post...
2 minutes ago, FreeThinker said:Make sure when reinstalling KSPIE, you remove Interstellar_Redist.dll from all Plugin folders (both WarpPlugin and PhtotonSailor) and it should only be present in the GameData folder
I also corrected the CKAN file for this change.
It was my understanding that KSP 1.12 changes on the DLL loading mechanism had solved this issue...
This is essentially what MM WatchDog prevents about MM, so what do you think to extend it a bit and check for multiple Interstellar_Redist.dll too and issue a warning if more than one is found? Ideally it should warn about its absence when it should be there too, but this is a bit more tricky (but not that much).
As it appears, this problem is going to stay...
-
When a problem is found, a page can be opened after quitting KSP.
Since I'm coding a WatchDog for Scale_Redist too, I will send "my" suffering users into TweakScale Forum Thread, as this is usually the fastest way to the user to get help. But this also means that the user will be freshly burnt by the problem, and not rarely a bit angry on you.Do you want that "yours" suffering users to be directed into Interstellar Forum Thread? Do you prefer them to be directed to some other page? Or just quit KSP and after the Fatality Box and call it a day?
-
-
About this post...
2 minutes ago, FreeThinker said:Make sure when reinstalling KSPIE, you remove Interstellar_Redist.dll from all Plugin folders (both WarpPlugin and PhtotonSailor) and it should only be present in the GameData folder
I also corrected the CKAN file for this change.
It was my understanding that KSP 1.12 changes on the DLL loading mechanism had solved this issue...
This is essentially what MM WatchDog prevents about MM, so what do you think to extend it a bit and check for multiple Interstellar_Redist.dll too and issue a warning if more than one is found? Ideally it should warn about its absence when it should be there too, but this is a bit more tricky (but not that much).
As it appears, this problem is going to stay...
-
Found the problem on your MM patch!
TL;DR: shove a :FINAL on the patch and it will work.Don't have an answer for this yet, it's working hours. I will try to explain that at night!
-
Humm... The bug is still there.
I recompiled MM with LOGSPAM, and noticed that with or wihout :FINAL, the patch enters in a infinite loop!
The difference is that, somehow, the :FINAL is not locking the thread.
I think that by not using :FINAL, the patch runs on the LEGACY phase, and whatever calls the LEGACY phase locks waiting for its finish, but don't lock FINAL.
-
-
Found the problem on your MM patch!
TL;DR: shove a :FINAL on the patch and it will work.Don't have an answer for this yet, it's working hours. I will try to explain that at night!
-
Hi all, answers to a few questions I've received.
Am I back? Tricky question, although having just viewed and edited around 960 cfg's and patches, all without flying into a ksp induced rage, is likely a very good sign. Mainly addressing tweakscale, MM and general cfg issues, further work is still required to prevent issues going forward.Will SM mods be back soon? (and all the associated questions, no links etc.) Its looking promising for some of the mods to return for 1.8.x, I doubt 1.8 is going to last long enough for me to get where I want to be, and knowing you lot, eager to make the jump I don't see any point in rushing out maintenance releases for 1.7.3
As it stands today what happens for KSP 2 depends entirely on what KSP 2 will require in the way of changes.
There are many things that it's easy to do, and many things that are neither easy or particularly fun to do, on the way to updating a mod, and should the changes require far too much of the latter and not enough of the former ( for example the idea of retexturing several hundred parts for pbr is nasty just thinking about it)I feel that they'll stay as KSP 1 mods.
That's not to say that there wouldn't be KSP 2 mods from SMI, because if the rumors are true our stuff could finally get some proper use
Cheers 0/ -
EVERYBODY! THE LIKES ARE BACK!
-
Life on the university campus:
It's been a while since I've shared one of these with you. Today's installment is about two students I have at a local community college where I adjunct teach. Both students scheduled appointments to come see me during office hours today. One scheduled at 9:00 and the other 10:00. Normally, my office hours do not officially start until 9:30 a.m., but because of the needs of students who want to see me, I often will be at the campus early. I truly believe in putting the needs of my students before my need of comfort.
9:00 a.m. I've been at the campus for about fifteen minutes. Noticed at 9:15, the student has not shown up for the scheduled meeting at 9:00. Check email - nothing. Check phone - no missed call. Simply no nothing! So, for the next thirty minutes what do I do? Surf the web and watch Family Feud videos of Steve Harvey. One by one, five students/former students trickle in to do the office hour hangout thing.
9:45 a.m. Student with the 10:00 appointment shows up early. So, I invite this student in my office and toss out the five "office toads" (term meaning students and former students who come to hang out in my office) out so the student can have my full and undivided attention. The student shows me the progress on their project paper for the course. We take care of some minor issues, student leaves at 10:30 satisfied and ready for the next step of the project paper. Still no word from my first appointment student.
10:50 a.m. Student from first appointment has still not shown up, no email no phone call. Walk across campus for my 11:00 class. I teach a 12:30 class right after that one. Finish my second class at 13:45 and head to parking lot where student from first appointment sprints to over where I am at, then laughs and says "I'm glad I caught you. I'm not a morning person and decided I wanted to see you this afternoon." So I asked the student why he made a 9:00 a.m. appointment and his answer was, "because I know you have early morning office hours. I'm not an early riser and since you don't have anything important to do, I'm here now..."
This student was shocked when I walked over to my car and proceeded to get in. He even asked, "can't we meet now? I'm here." My response was, simply No, I cannot meet you now. You missed your scheduled appointment at 9:00 this morning. I have off-campus obligations which need my full attention which is why I have early morning office hours. Please schedule a meeting next week since I won't be back on campus until Monday.
If he hadn't told me I didn't have anything to do and if he would have valued my time, this story might have had a happier ending for him.
Moral of this story
When you make an appointment with a professor you need to keep it. If you cannot keep it, get word to the professor explaining why you won't be there. Don't just not call them and act like it's not a big deal. A professor's time is as valuable as your time is to you so don't waste it. And after missing an appointment never tell that professor "since you don't have anything else to do..." because you will not like the response you get.
-
Moved out of the thread to prevent derailing it
21 minutes ago, Yeet_TheDinosaur said:Sure thing mate.
Wait....... Did I see you before?
Had your KSP crashed lately?
-
SNJO
Good Morning was playing with biplane last night and when I launch the plane it was sitting about a 1 KM above the runway. What is going. Thanks
-
(ok, nectroposting but the information is relevant)
It's a known problem with the FireSpitter landing gears. Use the stock landing gears to see if it helps.If not, you found a old part without a collision mesh (I have two on Impossible Innovations), but until the moment I found none on Firespitter.
-
-
Hi. I'm the dumb-SAS that had the terribly stupid idea of deleting the CurseForge project for StageRecovery.
I'm terribly sorry for even considering that stupidity, the CurseForge has 4 years of Project History, from the KSP 0.24.2 era! Would be sad to loose that.
I would like to kindly ask you to lock down the project instead, and edit the front page to state that new versions will be available on SpaceDock only. I think this is a better way.
I meant well, I was thinking on people relying on CurseForge only and that would be unattended of the change, but didn't though enough about the consequences.
Sorry for the mess, I should had known better.
-
Hi.
Your problem is related to RemoteTech . I'm 99% sure of it.
The MFI is just a "telephonist", it just redirects calls. Anything that goes wrong after the redirection, is due a problem on the end point. But since MFI is the one that made the redirection, it got itself involved on the stack dump.
As a side note, I kindly ask
to reach me here or by mail when something I wrote gots involved in the mess. It will save both of us some… Headaches.
-
Hi.
Your problem is related to RemoteTech . I'm 99% sure of it.
The MFI is just a "telephonist", it just redirects calls. Anything that goes wrong after the redirection, is due a problem on the end point. But since MFI is the one that made the redirection, it got itself involved on the stack dump.
As a side note, I kindly ask
to reach me here or by mail when something I wrote gots involved in the mess. It will save both of us some… Headaches.
-
Hummm…. I'm seeing the code that throwed the Exception, and it only runs when TweakScale finds a module called "ModuleFuelTanks" on the part.
I'm inclined to believe that TweakScale is innocent this time. My understanding about KSP is that it throws away orphaned data - so if the vessel was loaded and that part has a ModuleFuelTanks module on it, it must have a DLL that provides such module.
For my future reference: line 452:
https://github.com/net-lisias-kspu/TweakScale/blob/master/Source/Scale/Scale.cs
-
-
Hi.
Your problem is related to RemoteTech . I'm 99% sure of it.
The MFI is just a "telephonist", it just redirects calls. Anything that goes wrong after the redirection, is due a problem on the end point. But since MFI is the one that made the redirection, it got itself involved on the stack dump.
As a side note, I kindly ask
to reach me here or by mail when something I wrote gots involved in the mess. It will save both of us some… Headaches.
-
Hi.
Your problem is related to RemoteTech . I'm 99% sure of it.
The MFI is just a "telephonist", it just redirects calls. Anything that goes wrong after the redirection, is due a problem on the end point. But since MFI is the one that made the redirection, it got itself involved on the stack dump.
As a side note, I kindly ask
to reach me here or by mail when something I wrote gots involved in the mess. It will save both of us some… Headaches.
-
May I respectfully give you a bit of advise?
Feel free to ignore me if you don't agree.
That last guy on that "headache" of ours, he's right.
In my humble opinion, a better answer to him would be "Yes, now I know better - I will not do it again. But I meant good, and I though that because one of the errors happened on both forks, I could report it here so both of them could work together to find a solution."
Or some other niceness like that.
The sad fact is that a lot of people like to code, but very few like to handle users (in this case, you). It's like cleaning the room or doing the dishes - a pain in the SAS that everybody do, but try to do the least possible.
And since we also have family, hobbies, and games, every time we spend half a hour sorting a problem that in fact belongs to someone else, that's a half a hour that he could had used playing KSP himself.
As an example, I had played KSP less then 2 hours this week, half of them yesterday late night - since I'm committed to solve the most errors I can in the finite time window I have now due some idiosyncrasies on my country, I'm pushing me into this now as I know that I will not be able to do it later. But as soon as things start to work fine here, I will have the same problem too - so, I understand their point of view.
-
-
Hi.
Your problem is related to RemoteTech . I'm 99% sure of it.
The MFI is just a "telephonist", it just redirects calls. Anything that goes wrong after the redirection, is due a problem on the end point. But since MFI is the one that made the redirection, it got itself involved on the stack dump.
As a side note, I kindly ask
to reach me here or by mail when something I wrote gots involved in the mess. It will save both of us some… Headaches.
-
Search for all DLLs in your GameData.
Recently, I found a old ModuleManager forgotten inside a random folder from a Mod (the author probably dragged&dropped the file without knowing), and that was playing havoc with my installment.
That drove me crazy, as I was tweaking on the MM code, and I thought that I had made a mega-goof. It took me some days (and a lot of commits and reverse commits) until I realized that there were more than one MM on that !#$#!@$#@$#@ GameData! =P
-
-
Hi.
Your problem is related to RemoteTech . I'm 99% sure of it.
The MFI is just a "telephonist", it just redirects calls. Anything that goes wrong after the redirection, is due a problem on the end point. But since MFI is the one that made the redirection, it got itself involved on the stack dump.
As a side note, I kindly ask
to reach me here or by mail when something I wrote gots involved in the mess. It will save both of us some… Headaches.
-
Hi.
Your problem is related to RemoteTech . I'm 99% sure of it.
The MFI is just a "telephonist", it just redirects calls. Anything that goes wrong after the redirection, is due a problem on the end point. But since MFI is the one that made the redirection, it got itself involved on the stack dump.
As a side note, I kindly ask
to reach me here or by mail when something I wrote gots involved in the mess. It will save both of us some… Headaches.
-
I think it's another, concurrent mess.
Drag it's definitively involved in your problems with PersistentRotation and Atmosphere, and I found a Exception about it here:
Quote[LOG 19:58:56.420] DragCubeSystem: Creating drag cubes for part 'bluedog.Saturn.VFB.MissionModule'
[EXC 19:58:56.422] IndexOutOfRangeException: Array index is out of range.
ModuleAnimateGeneric.AssumeDragCubePosition (System.String name)
DragCubeSystem+<RenderDragCubes>c__Iterator1.MoveNext ()
UnityEngine.SetupCoroutine.InvokeMoveNext (IEnumerator enumerator, IntPtr returnValueAddress)
UnityEngine.MonoBehaviour:StartCoroutine(IEnumerator)
<RenderDragCubesCoroutine>c__Iterator0:MoveNext()
UnityEngine.MonoBehaviour:StartCoroutine(IEnumerator)
<SetupDragCubeCoroutine>c__Iterator2:MoveNext()
UnityEngine.MonoBehaviour:StartCoroutine(IEnumerator)
<CompileParts>c__Iterator1:MoveNext()
UnityEngine.SetupCoroutine:InvokeMoveNext(IEnumerator, IntPtr)Also, check your installment for TweakScale. Most mod I know depends from it in a way or another.
Quote[ERR 19:59:45.665] Cannot find a PartModule of typename 'TweakScale'
Interesting, I found this in the previous Log of yours.
Check the access permissions for this file and his directory path. Perhaps you have more of this? Would be a good idea to restore the file permissions from all files on your KSP!
Quote[EXC 20:04:24.160] UnauthorizedAccessException: Access to the path "D:\Kerbal Space Program\GameData\ResonantOrbitCalculator\PluginData\Cache\Sun.p
ng" is denied.
System.IO.FileStream..ctor (System.String path, FileMode mode, FileAccess access, FileShare share, Int32 bufferSize, Boolean anonymous, Fil
eOptions options)
System.IO.FileStream..ctor (System.String path, FileMode mode, FileAccess access, FileShare share, Int32 bufferSize)
System.IO.File.Create (System.String path, Int32 bufferSize)
System.IO.File.Create (System.String path)
System.IO.File.WriteAllBytes (System.String path, System.Byte[] bytes)
ResonantOrbitCalculator.PlanetSelector.Start ()(sigh).
Dude, there're NullReferenceException scattered on ALL YOUR LOG. From many different add-ons.
This is the reason I got fed-up and decided to fork and "fix" everything I use. =/
-
-
Hi.
Your problem is related to RemoteTech . I'm 99% sure of it.
The MFI is just a "telephonist", it just redirects calls. Anything that goes wrong after the redirection, is due a problem on the end point. But since MFI is the one that made the redirection, it got itself involved on the stack dump.
As a side note, I kindly ask
to reach me here or by mail when something I wrote gots involved in the mess. It will save both of us some… Headaches.
-
In time, not your fault.
I should had foresee this kind of problem, and the other guy could had be more civilized on their ways.
The Internet is wide and open - we will not receive from everybody the same levels of education and respect we are used on our private circles.
Some people, you just can't reach.
-
