On Dec 20, 2011, at 15:44 , Eugenio Furtado wrote: > Ultrasoft pseudopotentials in UPF format v.2.0.0 are affected by a > bug compromising their quality. Please regenerate pseudopotential > file for N
done (courtesy of Giuseppe Mattioli who, unlike myself, keeps track of generation input): http://www.fisica.uniud.it/~giannozz/public/N.star1s-pbe- rrkjus.UPF Please report any problem with it. The file on the web server will be updated soon (1-2 months), since we are moving anyway to a new web interface. > are the pseudo from the database of "general" purpose, I mean like > "lanl2dz" > or "sdd" are in gaussian or other MO packages? never heard about "lanl2dz" or "sdd" . The pseudopotentials "should be" suitable for general usage, i.e., unless you plan to do something special, they should be ok (note that the one you found to be buggy is for special purposes: calculation of X-ray spectra). Most of them have actually been used in serious (or pretended to be such) calculations. P. --- Paolo Giannozzi, IOM-Democritos and University of Udine, Italy "Nothing is impossible for the man who doesn't have to do it himself"
