This is necessary because of an internal bug on g++-4.2 on my mac

I removed the sentence saying that 2.2 will crash.

We should likely make it fail on travis-ci so that
we remember to add the documentation, but I cannot
do it know as one regtest basic/rt44b in v2.2
would fail.

Here we test for sum_hills using variables
that are sub-components. This triggers
the warning about non-registered components in FAKE,
that should be fixed someway

[makedoc]

[makedoc]

[makedoc]

This should show "master" instead of "2.3" in
online doc-master

This is c++11 so I added a check in configure.
However, I did not find any compiler not supporting it so
I could no properly test it.

Anyway, this should allow adding "explicit" in front of many
constructors that otherwise could be interpreted as conversion operators

Conflicts:
	src/multicolvar/MultiColvar.cpp

Conflicts:
	src/maketools/cppcheck

Includes statistics of (style) warnings.
Can be used to see if the number of warnings decreases.

[makedoc]

Conflicts:
	user-doc/bibliography.bib

[makedoc]

closes #147

[makedoc]

[makedoc]

closes #141

[makedoc]

[makedoc]

Before fix the master branch was installed
Now we explicitly ask for 1.69

Now $HOME/opt has the priority. This allows properly
overriding already installed programs.

[makedoc]

[makedoc]