Crzyrndm

No it's not just you, and yes it's probably something I should add. I just never get around to it... Could you log a github issue for it please, that way atleast it can't get buried again. Depends what you want to achieve with it, but the answer is probably So long as you're flying something in a manner even remotely similar to how a plane should fly, Pilot Assistant will handle it without issue (given correct settings ofcourse). Vertically, this includes things without wings that get lift mostly or entirely from engine thrust so long as the allowed AoA and craft control authority are high enough. Turning is another matter though... Just before switching to rocket mode I normally hand control to stock SAS since the steady "pitch up" is actually useful and PA doesn't handle attitudes directly. Re-entry I normally handle with SSAS so the AoA can be kept fairly consistent (switching to PA once AoA and VSpeed return to normal operational limits) Every time I have a survey contract I'm always so happy I got this working before they existed EDIT Something I forgot to mention at release time, v1.2+ has the ability to update the default presets ingame. It uses the current active PRESETS not the current active VALUES. If you want the current values to be the defaults, you need to make sure that you aren't using a default preset (which can't be updated directly), update the active preset, and then update the defaults. The extra step is a bit of insurance against accidentally turning the defaults into something insane

Version 1.17 Changed initialisation method to be more robust (completely independent of other mods now) Part name and title checks are now independent of case (AAA == aaa) Updated cargo bay checks to include the name Payload Bay (pull request from Gribbleshnibit8) Updated wing, strut, generator, and light checks with various fixes and additions (pull request from Wercho) Updated multicoupler check to be more robust. Parts similar to the stock staionhub with extra attach nodes on the sides of the part are no longer recognised as multi-couplers (ie. check is now for inline multicouplers) EDIT Forgot to mention, licensing has been changed. Icons and configs distributed with FE are under CC-BY-SA 4.0, plugin is covered by GPLv3.

Version 1.2.1 Resolved AGX incompatibility with keyboard controls - Thanks to Diazo for some help on this one Removed some log spam that found its way into 1.2 And since I forgot to post anything for 1.2 yesterday... Version 1.2 Added default update button Fixed default preset creation error with SSAS that was causing NAN control outputs

http://forum.kerbalspaceprogram.com/threads/99126-0-90-v1-1-Analog-Input-Fly-smoothly-with-keyboard-and-mouse

Hmm, something must have changed with a very recent version then. I'll see if I can replicate and get it sorted out with Diazo

Can you just try a test with only Pilot Assistant installed? If it works fine, these are the mods that I don't use, so the incompatibility is probably here one of these. If it doesn't, I'm slightly stumped as to what's happening. I may have to make a version with some extra logging to hunt it down. BahaSP MissionControllerEC PlanetShine QuickScroll ScienceContainers StageRecovery TouhouTorpedo TweakableEverything

dropbox, google drive, pastebin etc. all work quite well. This forum doesn't allow attachments so you have to upload it elsewhere

Oh yeah, theres a few leftover "don't interfere with me" functions that I've been identifying and cleaning up since they aren't needed any more All plugins and their install path are listed at the top of either log file. It's one of the reasons a full log is so important for support Just to clarify, the keybindings aren't responding as you would expect them to with respect to how Pilot Assistant uses them? (as listed in post 223 and the OP)

That will only promote users sticking to the same tried and true mods if they download any at all. How does that improve the OP's "issue"? (not that I see any issue raised by the OP. Creativity may be infinite, but inspiration and resources are finite)

Just to clear this up, 1.0.3 had a typo in the github release version number (marked as 1.03) and since that is no issue for anything other than CKAN I didn't realise or see it as an issue. Vysen identified the problem and since he linked me straight to it I *think* I fixed it by changing the version in the meta file for 1.0.3 to be correct. Anyway, I would suggest making use of githubs "Latest Release" tag if you can. That can avoid beta releases being treated as latest (use the pre-release flag) and typo's mucking up the automatic update system for mods that are added for w/e reason but aren't directly supported by the authors.

In future, please bug CKAN people about CKAN issues. I've fixed it this time since you gave me the link and details of the issue and it took all of 5 seconds (seriously, you could have fixed this yourself if you identified the issue. It would have been faster) CKANs problems are not my problem. I didn't put it there, and I don't particularly care for it being there.

Sounds sensible.

Ugh, I think I got confused by your confusion If I'm reading you/it right, the System.Object is being used as a generic boxing of whatever object you're actually requesting because you know what type you're getting, but overloading is either tedious or impossible (if a non-standard object can be sent through data?) EDIT This is just me being curious now. I've never run into the need for a pointer to a pointer before so I'm more wondering as to why it would be used.

The "ref" keyword means that data is being passed by reference not by value. In this case though, wouldn't it make more sense to use "out"? (it's ref, but it makes it explicit that the value of 'data' going in won't be used inside the function)

Ugh, small mistake on my part there thats probably going to come up. DO NOT CHANGE THE NAME OF THE PRESET It's saved separately but the association is by preset name, not which file it originated from. PS I'd recommend making the edits as a MM patch rather than directly. Makes for good insurance against things being overwritten.

In the Pilot Assistant directory there will be a "Defaults.cfg" (if you have launched a flight with Pilot Assistant installed. It's created on first run). That will have three presets in it, one for Pilot Assistant, and one each for the SAS modes. Replace the SASPreset that has "stock = true" with another SASPreset (easiest to copy out of the Presets.cfg) and the SAS for craft that do not have a craft preset should become what is set in the defaults. At the bottom of the defaults, there is also a craft preset. The true/false value in there is whether to use stock SAS or SSAS by default. After you use a craft and load/save any non-default preset, any time you return to a craft that has the same name (the only value that persists for multiple launches of the same vehicle and through tweaking parts on a vehicle) that preset will again be loaded for you (as it has been saved. If you load a preset and then do some tuning but don't update the saved preset, the original will be loaded, not your tuning)

Anything to do with CKAN is all on Kerbalstuff (apparently you even if you untick the CKAN integration it still gets added...) or whoever added it originally.

Something sounded strange about this, so I went and had a look at the source. tl;dr: It's not a partmodule and things here are heading way off course I am wondering why that section is split off from the part module, and why KSPAddon is used to initialise it rather than a constructor which looks to be more appropriate from my quick look. ie. // in the partmodule OnStart() thisClass test = new thisClass(/* initialisation parameters */) Add a constructor to the class that you are currently starting with KSPAddon and drop KSPAddon. The effect should be the same, but you get the code as part of the partmodule. If you want to enforce that only a single instance is running concurrently (may or may not be neccesary) then you can do things like use a Singleton variable instead of the constructor directly.

snippet from KER FlightEngineer.cs [COLOR=#333333][FONT=Consolas]ScienceUtil.GetExperimentBiome(vessel.mainBody, lat, long)[/FONT][/COLOR] I'm assuming that's a KSP class since I can't check right now and I couldn't find a ScienceUtil class in KER

Decrease max deflection of your front surface and/or making it smaller might help.

v1.1.3 Fixed incomplete presets causing errors Updated AoA override cfg to check for NEAR and increased default IClamps for stock users Reduced Throttle Increase/decrease sensitivity by 75% (was originally the same scale as the altitude controller...) Hard clamped outputs returning to KSP appropriately (ie. +/- 1 for control inputs) Keybindings list added to OP

Previously targetting 4.0+ only went wonky for some specific functionality (from memory, reflection results in fairly consistent breakages if 3.5 isn't targetted) and/or platforms (problems on Linux but not Windows). Compile against 3.5, it saves you many headaches if things do go wrong.

https://github.com/sarbian/ModuleManager/wiki/Module%20Manager%20Syntax#mod-dependency-checking MM allows :NEEDS[Mod1|Mod2|...]

Output clamps are absolute "this value will not be exceeded" things. I Clamps are an internal min/max on the integral sum that prevent issues caused by changes in setpoint (see: Integral Windup). The downside is that they also function as the zero error output clamp (if less than the output limits), so if it isn't high enough some offset will be required to achieve steady state.

OnDestroy should work for a part module. It gets called just before leaving the scene void OnDestroy() { GameEvents.onVesselGoOffRails.Remove(VesselUnpack); }