Hello,

I already noticed it to V. Acary but I want to point out that in the EulerOSI the case where an interaction has multiple linked DS does not correctly compute the sub matrices of K (tested on NonLinearRelation).