Corax

A backflip orbit has been discussed in this thread, also check the links from there. I think such an orbit might be possible if it weren\'t for the inaccuracies in KSP\'s physics model, and of course the lack of high-fidelity mission planning and flight computers, which may or may not be solved in upcoming versions or mods.

http://kerbalspaceprogram.com/forum/index.php?topic=8540.0

Maybe I wasn\'t clear enough, but I see this behaviour in other games, too. So far I haven\'t heard of framerate affecting shadows or antialiasing, only the other way around

Mainly just because I\'m busy developing and testing the telemetry plugin. I do like your mapper and especially the nice antenna part, and I think it\'s very good to have an integrated data collection/mapping solution The telemetry module, while it can collect elevation data, is primarily meant to make available as much other flight data as possible, and requires much more effort to process the data into something useful from the user\'s perspective. I\'ve mapped large parts of the Mun with ASL/AGL data; the main problem was that you basically had to orbit below the height of the highest peaks to gather any useful data. As you said, the old method was depending on LOD, but that was before PQS even existed. I stumbled on giucam\'s code in this thread, found it worked partially, and found how to make it work properly about the same time you all figured out the solution I wasn\'t aware of Mu\'s involvement, but it seems plausible he\'d know quite a bit about how PQS works Actually my data file is well over 8GB, with roughly half of it relevant to mapping data. I\'m using gnuplot to create maps, or any other data visualizations (thrust curves, atmospheric data, etc.), and while it does take its time, it hasn\'t let me down so far.

From the screenshots you posted, antialiasing doesn\'t seem to make much of a difference anyway... unless of course you had it disabled when you took them I\'ve got a Nvidia card, but I\'m also running Linux, and for me too AA doesn\'t seem to work properly, and/or produce texture artifacts when activated (not exclusive to KSP). I\'ve experienced this for at least three generations of Nvidia hardware now. It seems a bit unlikely that two competing hardware manufacturers have the same bugs in their drivers, but it is of course possible.

Just a question that comes to mind: did you start the receiver in the same VM instance as KSP was running in? Every virtual machine would have its own 'localhost', so it wouldn\'t work across different VMs or to the host OS. EDIT: I don\'t know how Parallels is configured out-of-the-box. You might have to set up routing in your VM to get access to other VMs or the host OS.

The problem with the previous implementation was that I only tried to connect to the first returned address; if that was invalid, no further attempts were made. Now as long as the DNS resolver returns at least one usable address, a connection will be established, so I don\'t think there\'s a need to check for specific invalid addresses. If for some reason the DNS resolver fails to return a valid address, explicitly specifying a known working IP address in the part.cfg should work anyway (dotted quad, i.e. '127.0.0.1' for localhost, or a routable address if you want to connect to another machine), even with the current version of the plugin.

Thanks again. I have no idea about how a VM might influence DNS resolution, but the new code will try to connect to the next address in the list until it succeeds or runs out of options.

With the last released version (2012-04-08), you can get ASL and AGL, from which you can calculate terrain elevation, as long as you stay low enough. The current version (waiting for 0.15 to release) will use giucam\'s method of getting terrain elevation and so will work (is working) from higher altitudes as well.

I\'d reckon there are quite a few duplicates in there, I haven\'t done any cleanup yet. Also I\'m using the telemetry plugin, so no extra-wide scan paths. I\'ve put up a map there, I\'m still gathering more data to refine the map resolution.

Looking at about 4 gigabytes and growing (and yes, I mean milliards), that\'s a plain no

Most likely because it isn\'t

Because I wanted the way of addressing the receiver as flexible as possible: the string _ipAddress might already contain a valid IP address ('123.45.67.89'), but it could also be a DNS name ('myPC.mydomain.whatever'), or 'localhost', or an invalid address altogether. Obviously I didn\'t test thoroughly enough... but it should be fixed in the next version ('should' because I don\'t have a Mac for testing). In the meantime, thanks for the workaround

Here\'s a map and elevated view of Kerbin from data I\'ve been collecting over the last two weeks or so, while testing the latest version of the plugin (to be released shortly after 0.15 is out and reasonably stable). (click for full-size) BTW still experimenting with the gnuplot palette mapping, if anybody knows a good mix for topographic maps, I\'m game...

youtube? ??? I thought it was common sense. DUH. I guess that was some kind of a hint to put a link to the videos in the OP.

But where does it say so? The Parachutes class doesn\'t have a member of that name, and you haven\'t declared it anywhere...

And where does it say so? (Apart from that, I think it ought to be float)

Right. And of what type?

Think about what exactly 'deflectionLiftCoeff' is.

You might want to work through a few basic tutorials of C# first before trying to code a plugin... Basically, every instruction needs to be terminated with a semicolon. On a closer look, your code is missing even more than one.

This: ';'

Code does not 'refuse to compile' with no reason. Look closely and you\'ll see you\'re missing a semicolon.

Another free, open-source flight sim is Flight Gear. I am no pilot, so I can\'t really judge how realistic its physics model is, but I think it flies quite well. The graphics aren\'t that polished yet, but Scenery covers the whole world, and there are tons of common as well as more exotic aircraft to fly. Also it runs on Linux and Macs, too. And being open-source, and having an open protocol, it is pretty flexible as well.

Everything else aside, no, for me that isn\'t enough by far. Where I live, there\'s a principle called 'innocent until proven guilty'. When someone is accused of a crime, he may be taken in, but that is not any proof of guilt. Even if the evidence is against him, as long as there is no proof, the accused is to be seen as innocent. Various political groups are constantly trying to undermine these principles (no, I\'m no conspiracy theorist, but I see with concern how stating things like 'these forums are not a democracy, you have no rights here' has a subversive effect in line with that process).It is my view that basic rights, such as the principle of innocence, or freedom of speech for that matter, must be valid here, too. If someone thinks of that as off-topic, I think this is exactly the right place and time. Also, if you have doubts about something, either confront the person openly XOR in private mail, XOR report it to the mods, but then stick to the way you have chosen; unless of course there is no progression. Keep it in the open, XOR keep it in private mail, XOR let the moderators deal with it. But don\'t go around announcing your deeds as if you had rescued the world. Especially if nothing has been proven, and all your evidence are a comment and a debatable representation of a number, and all your proof is that the thread has been hidden from view. I myself have (hand-)written comments strikingly similar to this when I started the first iteration of my plugin, simply to remind me where to look for the stuff I\'d need. EDIT: taken out uncalled-for accusation

I haven\'t seen Cepheis code, but I\'d like to know on what you base your accusations of 'decompiling' - it seems you are using this word quite a lot. This is what is publicly visible of the Liquid engine class by reflection without any need for decompiling. All of this info is (surprise, surprise) contained within Assembly-CSharp.dll: public class LiquidEngine : Part { // Fields public float minThrust; public float maxThrust; public float fuelConsumption; public float heatProduction; public float gimbalRange; public float gimbalAngleH; public float gimbalAngleV; public Vector3 thrustVector; protected Vector3 thrVector; public bool thrustVectoringCapable; protected float thrust; protected Part fuelSource; protected Transform gimbal; protected Quaternion initRot; // Constructors public LiquidEngine (); // Methods protected virtual void onPartStart (); protected virtual void onFlightStart (); protected virtual void onCtrlUpd (FlightCtrlState ctrlState); protected virtual void onDecouple (float breakForce); protected virtual bool onPartActivate (); protected virtual void onPartDeactivate (); protected virtual void onActiveFixedUpdate (); public virtual void OnDrawStats (); } EDIT: Ok, so now I have seen Cephei\'s code (the web never forgets...), and all I can find is this 'obvious comment at the beginning' // Type: LiquidEngine // Assembly: Assembly-CSharp, Version=1.0.0.0, Culture=neutral, PublicKeyToken=null // Assembly location: D:\Games\KSP\KSP_Data\Managed\Assembly-CSharp.dll using System; using System.Collections.Generic; using UnityEngine; namespace KSPPartTest { public class IonEngine : Part [snip] which doesn\'t really hint at any illegal activity, if you don\'t count using backslashes for file paths. Regarding that ominous 0.699999988079071, that is a close representation of 0.7, as the exact value of 0.7 can not be represented using IEEE 754 double-precision floating-point numbers; again I don\'t see how this would give away any 'decompiling'. I\'d suppose a decompiler would be even more precise, as this is still not the exact value of 0x3FE6666666666666 if you use 64bit-numbers, even if represented as a 32-bit float it would have more significant digits than that. I\'m not omniscient and certainly not infallible, but I can see nothing in Cephei\'s code that would back up your accusation.