Nereid

Just add a line 9.9999 L+ Kerbal Name 9.9999 M+ Kerbal Name for each mission. Replace 9.9999 with any timestamp you prefer (it determines the length of the mission) and keep all timestamps in ascending order. But a question for all: Any issues with 64 bit? I have replaced my development installation with 32 bit, because on 64 bit every second or third save, it gets corrupted (game crashes on load). I don't know if FF (my current, not yet released version) or any other mod is causing this or its just vanilla behavior on 64 bit.

Nice to hear this. Because 64 bit has still its issues, ATM is still the most important mod. Its the only way to run a bit more than just vanilla.

NanoGauges 0.5.10 is online New gauge: Q - dynamic pressure (FAR required)

First, just as some kind of anecdote: a log of a size of about 1MB isn't that massive. Normal applications are often producing several of them per day, per process. But they are not running at log level TRACE of course (if they would, they would produce several gigabytes of log outputs!) But anyway... As far as I can see: FF is installed correctly there is no mod installed which alters the kerbol system Trying to reproduce it myself... which I can't. All is working flawless for me. And to be honest: FF doesn't differ in its behavior between career and sandbox mode. So I really doubt that this is caused by FF. But this is quite strange: [LOG 00:45:00.408] FF: adding toolbar buttons [EXC 00:45:00.409] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.420] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.475] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.612] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.624] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.637] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.647] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.655] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.663] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.670] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.678] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.692] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.709] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.726] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.742] NullReferenceException: Object reference not set to an instance of an object [LOG 00:45:00.894] FF: destroying Final Frontier But I still have no clue, what is causing this... Are these exceptions still present in your current log?

But its still not working? EDIT: Add a folder for texture Pershina. To do this yourself, take a look in the KSP.log after each step and search for texture Nereid/FinalFrontier/Ribbons/XXXXXX/SphereOfInfluence not found XXXXXX is the name of the celestial body that is missing. If you do this, the ordering of the ribbons is still messed up, but that's a cosmetic issue.

You could do with the ribbon graphics anything you want. The names of the three suns have to be changed in the game. Adding the folders alone would not work. There should never be a celestial body with the name of another body in the game. Reason for this: Any ribbon is based on an achievement in FF. They are created dynamically for the celestial bodies (good news for addons that create new ones; but thats only half part of the story). Each achievement is an Activity and each Activity has a code. All codes must be unique, but if some celestial bodies are named identical, the codes were identical and added twice to the pool. This will cause an exception and KSP wont start anymore. I need this behavior myself, because this way I notice by a test, if I have done something wrong with the codes. And if they get messed up, all ribbons for players may get messed up.

Has this ever been working with FF? Maybe a quick fix, that could (or could not) work: Just copy some of the other folders (like Duna for example) and rename it to Manai. Do this for all celestial bodies. But additionally you have to make sure that there is only one Sun. Rename the other two to Sun2, Sun3,... and create folders for those, too. And please modders... keep it simple. Don't add celestial bodies with the same names. This will make things way more complicated than it should be.

Thanks for the log. Investigating... The toolbar could not cause this as long as you use a recent version. The FAR optional file is that: optional. That couldn't cause this, too. EDIT: BTW: Are you using mods that alter the kerbol system? The line [WRN 10:21:27.576] FF: no base prestige for celestial body Manai shows that there is a celestial body Manai and I'm quite sure it is not in vanilla. This causes: [ERR 10:21:27.578] FF: texture Nereid/FinalFrontier/Ribbons/Manai/SphereOfInfluence not found [EXC 10:21:27.579] NullReferenceException: Object reference not set to an instance of an object And due to a bug (introduced by me in some earlier versions) a null reference is accessed whenever a texture is not found. There is another log output that confuses me a lot more: [ERR 10:22:47.949] FF: not unique code 'I:Sun' in Nereid.FinalFrontier.ActivityPool FF tries to register an activity twice and that should never happen. I will have to check if I have introduced this bug or if it is caused by other things. EDIT2: I think this is caused by the mods you are using. My installation only has a single Sun. I think yours has at least three (all with the same name)... But first: remove all mods that alter the kerbal solar system. Currently FF won't work with those mods. Its still on my list and a bit more complictaed to implement. Ok, this is solved I think. I'm taking a break now and then I will take a look in the log from MarcRan17 .

Can you please provide the KSP.log? There is a nullpointer exception. Maybe this is caused by a missing texture. Note for me: make FF more robust in that part of the code. We already had some strange effects like zip archive not extracted completely (i.e. ribbon graphics were missing) EDIT: I knew about those symtoms. I usually have them myself if I have messed something completely up. Maye this is the case here, maybe it caused by other things. I need a few hours to get this solved. EDIT2: Installed the archive directly and all is working fine for me. So I do not have messed all up and zipped the wrong dll. Unfortunately this is not the best case, because its makes it harder to reproduce...

You made my day

Yeah, thanks. That IS indeed helpful. I can't really analyze this just at this moment, but a few observations: The log level FF is set to detail or trace. This causes the amount of logging. But for debugging, this is good news. The line "[LOG 00:46:26.714] FF: no element for code M10 found in pool (Nereid.FinalFrontier.ActionPool)" tells, that there is a code in the halloffame.ksp, that FF dont know. But this should not break the game, its just gets ignored. there are some NullReferenceExceptions that CAN be caused by FF. Maybe FF is not causing this, but I will check if there is a bug just after loading the hall of fame. EDIT2: I was wrong in point 2. One of the last changes in FF was the implementation of different pools for ribbons/actions. And M10 is a ribbon not an action, so its quite normal not to be found in this pool (FF checks anyway, because it doesn't know this until it gets it in the other pool). So this is not your fault. Conclusion: Do you have edited the halloffame.ksp? If yes, maybe there are some lines incorrect (it seems so). Anyway this should not cause any malfunctions of KSP or such strange effects in the GUI. But the NullReferenceExceptions can cause this (and maybe some new bug made FF less robust against such errors). EDIT: Yes, the log level of FF is trace: [LOG 00:45:36.379] FF: log level is TRACE But again, for debugging or reporting a bug, this is more helpful than INFO. I really wish all user would give THAT amount of information without beeing asked. EDIT: Did you change the log level intentionaly? If not: Anybody else with log level TRACE by default? EDIT: What the hell is happening here? [LOG 00:45:00.408] FF: adding toolbar buttons [EXC 00:45:00.409] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.420] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.475] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.612] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.624] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.637] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.647] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.655] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.663] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.670] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.678] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.692] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.709] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.726] NullReferenceException: Object reference not set to an instance of an object [EXC 00:45:00.742] NullReferenceException: Object reference not set to an instance of an object [LOG 00:45:00.894] FF: destroying Final Frontier

You said that people are confused. They aren't. Its clear what to do. Just extract the folder. That's not true. I'm not breaking the purpose of a filesystem. I'm using the filesystem as it IS supposed to do. You are completely wrong in this case. EDIT: Is it really that hard for a user to use the search system provided by the file explorer? A quick search gives the result in seconds. I will NOT discuss with you that way. I have told you, why I have made that decision. I won't change it. Respect this, or go away. Nothing more to say.

Ah, thank you. Maybe we could get this to work. I was also thinking about calculating Q myself, because it seems to be not that difficult. I will try your modified class later this day, but I'm currently at work. EDIT: I will revert the changes GetMachNumber method, because I need the call with parameters in FF and maybe in NG later (horizontal mach number).

Ok, I will try to analyze this. I still had no time to play career or do anything else than playing around with sandbox a bit. But currently I don't believe this is caused by FF. Maybe a strange KSP/Unity-Bug. There is only one change in the new FF that eats some GUI events to prevent a click-through in the GUI windows. Unfornately this doesn't work for buttons anyway, but maybe this have some strange sideeffects, too. EDIT: Which version of KSP are you using (32 bit, 63bit)? Wich version of FF? Overlooked the first line in your posting. Answer is x64 0.24,

what the hell... trim? I've been stumbling over this in the KSP-API while I was creating NanoGauges. Then I was looking for trim controls didn't found any and thought this was reserved for later use. There are trim controls in KSP? What keys?

If you want to integrate this plugin with FF, I could provide you with an API that gets all necessary data and would work even without FF installed. Just write me a mail and give me a few days.

I have tried to read this value. But at the moment my knowledge of reflection in C# is quite low. I can't get to this value, because it's private field in an instance (well, the access restriction is not the problem, but to access the instance is). Maybe at a later time I can access this value. but a better solution would be some kind of public access in FAR.

I could restrict this to parts with heatshields of course. This could be done.

The next version of FF won't require the FARAdapter.dll anymore.

Do you mean http://forum.kerbalspaceprogram.com/threads/84409-0-24-0-NanoGauges-tiny-ana1og-gauges-for-kerbalnauts-0-5-9?p=1286553&viewfull=1#post1286553? That's 0.5.9 for NanoGauges.

I will have to check the source of DE for that... EDIT: if I understand the source correctly, DE is managing the temperature for each part independently. So yes, it's possible to add a gauge, but this would have a major drawback: All parts have to be scanned over and over again. I already have solved problems that were caused by this and I don't want to introduce them again.

Well, your right in some way, but wrong in some other. Developing software is an expensive task. Modern software depends on frameworks and is very complex compared to software ten or more years ago. So you have to make a decision: get the most out of the hardware and optimizing ram usage, performance and so on or just tell the user to add a few gigs more. Ok, what is the price for let's say 8 GB ram? A few hundred to a few thousand dollar (it depends on the hardware). And what does it cost to reduce the memory usage of an application? Let us assume two developer spending three months of time for it. Each of them has a salary of 80,000 dollar per year. So optimizing costs 40,000 dollar and that's not including the possibility of new bugs that have to be solved later. And unfortunately those developers were needed for other tasks and and you can't hold your deadlines if you spend time for optimizing... another 50,000 dollar spent for external support, additional meetings with the customer, lawyers,... Do you still think, it would be best to use less ram? I hope no. So it's not technical ignorance. It's just economic efficiency.

Consider it done. I will upload 0.5.9 in a few minutes.

Ah, ok. I will see what I can do. I have found a solution. The new Adapter class handles this with reflection. Yes, reflection is slow, but I have done some performance tests and calling the method that calculates the mach number using reflection 100,000 times costs about 100 ms. So calling it each frame (about 100 times a second), costs 0.1 ms per second. I think we could spend this easily. The mach number gauge is implemented in 0.5.8. EDIT: What do you mean by "Q-gauge"?

That's the reason why I will never include this plugin in my own mods - until is has some kind of "opt in scheme". Modders have to respect the privacy of users. Personally even some kind of automatic update or any check if there is an update available should at least be optional. So: full ack.