Hi Bjorn,
Using windows 10 and FEFLOW6.2/64 bits.
I'm having problems with the pointers from the MNW functions. The plugin have been using the pointer info before it is populated in my plugin, which makes me think that the compiler optimization has somehow changed the sequence of commands. As a result, lots of NaN's, negative node indexes and strange numbers appears throughout the model run. When I use it with the debug version, it works perfectly without any problem.