> Sergio has ported libfam0c102 to add a provides for libfam0 which is missing. Hmm? Surely this is not correct? Unless I'm misunderstanding the situation terribly, something compiled against the old gcc2 libfam0 (and thus which depends: libfam0) cannot link against the new gcc3 libfam0c102 (different ABI), and so libfam0c102 is telling a lie if it provides: libfam0? b.