-
GILLES Sebastien authored
#820 Fix a bug introduced yesterday: the matrix used in dH was not the right one (Dirichlet contribution was still there; it didn't appear on numerical results of first iteration but could have been tricky to find in later ones).
d85479ab