Cr convergence, better inpgen defaults/ warnings for magnetic systems
Here is a unit cell of Cr in some structure from the ICSD.
Fleur takes the inpgen defaults (inp.xml below). But the system does not converge at all. distances from 50 to 700 me/bohr**3
I do not know what the issue exactly is here. Maybe there should be better initial magnetization defaults for inpgen and/or if this is not possible at least a fleur warning for the user that he should set some starting point. I leave this here for investigation.