boost::shared_ptr<>::value_type has been removed.

Artistic Style Version 2.02.1

With the following options:
astyle \
    --style=allman \
    \
    --indent=spaces=4  \
    --indent-switches \
    --indent-col1-comments \
    \
    --pad-oper \
    --pad-paren-out \
    --pad-header \
    \
    --convert-tabs \
    --break-closing-brackets \
    --add-brackets \
    --align-pointer=type \
    --max-instatement-indent=80 \
    --preserve-date \
    --lineend=linux \
    \
    --suffix=none \
    --verbose \
    --recursive "*.*pp"

Artistic Style Version 2.02.1

With the following options:
astyle \
    --style=allman \
    \
    --indent=spaces=4  \
    --indent-switches \
    --indent-col1-comments \
    \
    --pad-oper \
    --pad-paren-out \
    --pad-header \
    \
    --convert-tabs \
    --break-closing-brackets \
    --add-brackets \
    --align-pointer=type \
    --max-instatement-indent=80 \
    --preserve-date \
    --lineend=linux \
    \
    --suffix=none \
    --verbose \
    --recursive "*.*pp"

now it is possible to instantiate a FSIMonolithic directly, instead of
an FSIOperator. This will smooth the forthcoming splitting of segregated
and monolithic.

In addition a couple of methods to export just pressure and velocity has
been added.

This generality is particularly useful for the Multiscale framework.

working for monolithicGI.

Right now the library compiles all the modules of the library
and all the testsuite runs and it's ok

updated the test.

set the conservativeFormulation flag to FALSE as default value

Due to some legacy in some parts it is not yet removed but just
deprecated.

The restart work both for domainVelImplicit true and false.

The test monolithic has been by default to fail and, if it works, the state variable
returnVariable is set on ok.

make the new framework usable by the Multiscale

changes in the FSI solvers, deleting commented parts, uèdating the conservative/non conservative switch

the solution vector is stored in the TimeAdvance class, which should be the owner of the vector, responsible of initializing and destroying it.
Avoiding the copies of the solution vector in the solver classes also helps the initialization.
Fixing a bug in the time distretization (thanks to Fabio and Simone), the continuity of the velocity was discretized in an inconsistent way 
for order greater than one.

fixed bugs for the TimeAdvance
fixed the non linear Venant Kirchhoff material. Now it's working also in FSI monolithic

*fix of the extrapolations of the ALE map (file FSIMonolithicGI.cpp)
*the FSIMonolithicGE file has still to be modified
*the possibility of plugging a nonlinear structure has been implemented
*the initialization of the TimeAdvance is fixed

I updated the TimeAdvance for the monolithic solver, now it seems to be working, though it is preliminar and I still did not test it properly.
I did also several modifications to your files, so check them before updating. Next steps is the implementation of the analytical benchmark
so that we can verify the orders of convergence.

second order initialization under development

* Compatibility changes with the modifications in FSIOperator and
FSIMonolithic
* Few cleaning + removing solveJac (the same method is in the base
class)

small addictions/corrections/indentation

hopefully solved the problem with the restart of FSI GI simulations, now the solution of the
restarted and non-restarted simulations perfectly match

hopefully solved the problem with the restart of FSI GI simulations, now the solution of the
restarted and non-restarted simulations perfectly match