The most severe change probably is the removal of the AutoDiff
debugging helper functions which were useful from within a debugger
but unfortunately had to rely on a presumed linker bug in order not to
be removed in the final binary.
Also, some private attributes were unused. These have been removed and
the constructors of their respective classes have been adapted. Once
their intended functionality is actually implemented, they should be
brought back on an as-needed basis.
Thanks to @bska for the review!
This should simplify some uses of the autodiff code. The internals
have been changed to allow for objects to have an empty vector of
Jacobians, always treating that object as a constant.
CLang and recent GCC warn about the "typedef" 'OneColInt' in
AutoDiffHelpers.hpp being unused. Similarly, GCC warns about unused
parameters in various place at level "-Wunused". This change-set
either removes ('OneColInt') or suppresses those messages.
There is some code in place now to create wells for the no-deck case,
but since it does not work correctly yet, the simulator intercepts this
and throws.
make all non-implementation headers includable without
preconditions. Also, this removes the GravityColumnSolver.hpp file,
because it tried to include a non-existing file and it was thus unused.
for some of these files this is needed to make to keep it compiling
after the next patch because the new ErrorMacros.hpp file will no
longer implicitly includes <iostream>. for the remaining files it is
just good style.
While at it, the includes for most of these files have been ordered in
order of decreasing abstraction level.
The state that is passed to the simulator object is directly
accessible without any encapsulation towards the client. After
the notification callback was introduced, this allows the client
to observe the state in the middle of a simulation.
However, it may be that the simulator has some internal state
which is not reflected in the state object because there is a
cost associated by flushing it into the TwophaseState format.
The notification is called back on every timestep, not just the
ones that will do reporting. It may even be that reporting is
done dynamically and is not known at the time of setup. (It is
more like a condition variable).
Consequently, flushing the state in every timestep is a bad
idea. This patch sets up a new method sync() which it is expected
that the notification will call if it needs the state for
reporting purposes.
Currently it is a no-op. It just establishes a protocol that
other, compatible implementations can also use.
