ИСТИНА

Current understanding of the final stage of gravitational collapse in General Relativity (GR) is based on a solid theoretical foundation involving singularity, no-hair and uniqueness theorems which suggest Kerr metric as a base for astrophysical modeling. This picture, however, is susceptible both to possible modifications of GR (not discussed here), and to changes in key assumptions such as cosmic censorship (CC) and chronology protection (CP) which are not strictly proven or proved in a model-dependent way. In anticipation of images form the Event Horizon Telescope one is interested in alternatives to Kerr in order to parametrize possible deviations from the standard picture. Several proposals were made in this direction such as "bumpy BH" or "modified Kerr" which suffer from naked singularities (NS) and/or closed time-like curves (CTC) regarded as pathologies. Based on our recent results obtained jointly with Gérard Clément we discuss to which extent NSs and CTCs should be vewed indeed as unacceptable features. We revisit Taub-NUT solution without periodic time identification and show that Misner string (MS) singularity, contrary to expectations, is transparent for geodesics and, after extension at the horizon uncomplicated with Hausdorf problem, is geodesically complete. We also show that CTCs encountered in the vicinity of MS are not geodesics and are harmless for a distant observer. We show that Reisnner-Nordstrom-NUT solution with supercritical charge becomes a wormhole which does not require violation of energy conditions, the only price being the MS itself surrounded by CTCs. Meanwhile, the geodesics do not stop at MS, while the existence of CTCs is likely to be observationally harmless. We analyse distributional Ricci tensor on MS and discuss possible matter sources for it. Finally we present an analysis of the Clément solution (1998) giving some rotating and charged generalization of the gamma-metric. We show that this solution presents the system of two dyonic corotating NUTty black holes connected by a charged and magnetized spinning string acting also as MS and the Dirac string. The solution is free from strong curvature singularities and contains only a mild distributional singularity on the string. We argue that GR solutions with mild NSs should not be rejected as patological and are worth to be looked closer in astrophysical modeling.