Follow-up Comment #6, bug #21080 (project relax): I can confirm this bug has been solved. Today I can run the analysis on my windows machine. And I tried 4 different folder naming and results placement. On local machine, over ssh, at attached network drive. And it is all running nicely. Yesterday I also had problems running an analysis with spins deselected for a clustering/global fit analysis. See: http://article.gmane.org/gmane.science.nmr.relax.devel/4455 Something with number of parameters not matching, and the errors were reported for the spins which were de-selected. But today I can't reproduce the bug. So I thank you for quick fix ! I think it was either me (or my linux comp) who was a little confused with the intermediate update on the source code. Thanks again. :-) _______________________________________________________ Reply to this item at: <http://gna.org/bugs/?21080> _______________________________________________ Message sent via/by Gna! http://gna.org/