Minor updates to the ANARCHYDU scheme

See merge request !961

Add a fake initialisation of the equation of state global objects

Closes #626

See merge request !959

Do not overflow the counter of extra particles by using a size_t where appropriate.

Closes #627

See merge request !960

Changed the confusing values of the min/max energy fraction of the EAGLE feedback in the parameter example file.

Do not dither the extra particles for star formation.

Closes #625

See merge request !958

Fully implement the timestep limiter (non MPI)

Closes #196

See merge request !956

Change the default primoridal Helium fraction to the value from WMAP7 that was assumed in all the EAGLE and OWLS runs.

In the black hole reposition code, calculate a delta position rather than an absolute position. This correctly handles the periodic boundary conditions. Also update the x_diff of the BH particles.

Reset the reposition variables once the reposition is done. Avoids problems when there are multiple drifts.

Change the EAGLE SNIa DTD to a properly normalized power-law of slope -1

See merge request !954

Only check the calculation of the minimal neighbour time-bin if we are not using vectorization. This feature is not used in the Gadget-SPH SIMD code.

Revert "When dumping a restart file at the end of a run, only drift the particles if no other form of i/o has taken place this step to avoid a drift towards the past."

This reverts commit d93c56e9.

Do *NOT* use the default values as optional parameters when re-reading the space global variables when resuming a simulation!

When dumping a restart file at the end of a run, only drift the particles if no other form of i/o has taken place this step to avoid a drift towards the past.