Fix Windows-specific bug due to not allocation QCoreApplication for the console-mode running of the mzML/db conversion.

the isotopic cluster.

Set decimals to 12 for the double spin boxes so that the slices are correctly computed without missing border points.

Refactor code to make the data loading feedback similar to all to other feedback routines in mass spectral computations.

The progress feedback is fully informational. Finished deporting to QThread the tic intensity signal computation for drift spectrum.

Fix bug due to not testing for infinity of the mobility drift time value. That is odd, because I never checked for isInf() but only for isNaN() and that used to be enough. A compiler change?

Incorporate all the modifications to the code to move all blocking
computations to QThread. Note that the code using OMP is still there,
but it runs inside a single QThread-based thread.

Greatly reworked the parsing of mzML/mzXML and sqlite3 db format mass spec files for making it possible to run the mass spec data integrator in an isolated thread.

Implemented the QThread-based threading for integrations to mass spectrum and to drift spectrum. Major rework of the coloring of the plot widgets so that multiple simultaneous integrations might be started happily with no mixing.

Remove crufty MassDataIntegrator member data since the integrator is now allocated at each new integration invokation.

Now the mass data integrations to MZ are totally pararallel and can be triggered multiply at the same time.

The MassDataIntegrator class is now a worker class that can work in a QThread. But the QThread is dynamically allocated in the MassSpecWnd and it should be a member datum. Going to do that change.

Fixed problem with the data relationer. But the cancel push button emits a signal that is not relayed to the integrator prior to the thread finished.

first worker thread implementation that goes to the mass spectrum in another thread still using openmp for the combination. Only proof of concept code at the moment.