Initialisation of running engines confusing

[seanmcleod]

I have an aircraft with 2 engines and I wanted both engines to be started by the initialization file. Looking at FGInitialCondition.h I saw the following comment:

// - running (-1 for all engines, 0 for no engines, 1 ... n for specific engines)

So I added a running element to my initialization file.

<running> -1 </running>

However neither of the engines are started.

I then also noticed that the code in FGInitialCondition.cpp looks for multiple running elements, and based on the comment above - 1 … n for specific engines I tried the following based on the comment regarding 0 and 1.

<running> 1 </running>
<running> 2 </running>

However that only results in engine 1 being started and engine 0 isn't started. Engines are 0 based, from FGFDMExec.cpp

  for (unsigned int n=0; n < propulsion->GetNumEngines(); ++n) {
    if (IC->IsEngineRunning(n)) {
      try {
        propulsion->InitRunning(n);
      } catch (const string& str) {

So looking at the code in FGInitialCondition.cpp.

  // Check to see if any engines are specified to be initialized in a running state
  Element* running_elements = document->FindElement("running");
  while (running_elements) {
    enginesRunning |= 1 << int(running_elements->GetDataAsNumber());
    running_elements = document->FindNextElement("running");
  }

I then tried the following which now does start both engines (0 and 1).

<running> 0 </running>
<running> 1 </running>

So assuming that the engine indices to the running elements should be 0 based then the comment in FGInitialCondition.h needs to be updated.

// 0 for no engines, 1 ... n for specific engines)

And the code needs to be fixed to properly handle the -1 option.

I'm happy to make the fixes, but before I do I just wanted to double-check whether the engine indices in the running elements should be 0 or 1 based.

[bcoconni]

It is very likely that I am the culprit of this (due to the commit 66ceba9 that does not handle the case where -1 is specified).

Not sure that the case 0 ever worked however. After all, if one does not want the engine started then he/she would not specify any <running> element ?

So yes please fix this and sorry about that.

[seanmcleod]

Doesn't look like the code before your commit handled the -1 case either. I'll submit a pull request to fix the handling of the -1 option and update the comment that engine values go from 0 to n-1.

[bcoconni]

I just pushed the PR #103.