Related to #175

[makedoc]

[makedoc]

closes #158

Since plumed 2.2 creates a directory named src/lib/install
upon install, there could be ambiguity in the install target
in plumed 2.1 when switching branches without cleaning.
Now the target is PHONY, and should give no issue.

[makedoc]

MULTICOLVARDENS is not present in 2.1

Same as #171 but using C++ standard include file <ctime>

Closes #171

[makedoc]

Date in changelog fixed on release

[makedoc]

[makedoc]

a full explanation was already there.
I added a link to it

When you type [makedoc] the developer manual is updated online

[makedoc]

Some more automatization of relase:
1. Version number is guessed based on current branch
2. Date in changelog is automatically set

[makedoc]

[makedoc]

[makedoc]

[makedoc]

Related to #165

[makedoc]

[makedoc]

Related to #168

This bug was introduced by me in 5e3aaf3a
The barrier on multi replicas was due to overcaution and
is only necessary when using WALKERS_MPI

The kernels were normalised incorrectly and their derivative when the
covariance matrix was non-diagonal was computed wrongly.  This check
affects a number of regtests.  analysis/rt0 changes beecause the kernels
were normalised incorrectly.  basic/rt44, basic/rt48, basic/rt49, basic/rt52
change because the derivatives output in the fes file have changed sign

Lattice reduction is used in plumed to convert box to its
"least skewed" form. I changed a bit the threashold since it
was giving problems with high optimization flags.
Additionally, I added some warning when too many iterations
are performed (which indicates a numerical issue)

reduced precision to make them work with stronger optimizations

[makedoc]

It seems that sometime on travis a regtest with mw-mpi crashes.

I think this is because when using multiple processors per walker
there is a problem in the opening of the old file (for read) and
its backup (for write). Adding this barrier seems to make it more robust