- Altered Nsf_NextHashEntry() to re-init hSrchPtr when the number of expected entries differs from the number of real entries. This fixes a bug that Michael Aram detected, that happens when multiple hash buckets exist, but on deletion of an hash entries triggers some other deleted from the same hash table. - extended regression test.
- The fix of yesterdays subst reference counting bug triggered an other problem: If the last arg was "args", the flags array for checking the need for decr of refcounts was not extended. There are multiple possible fixes, i have just extended the flags array for now as well. - When profiling is activated, perform now a more sane shutdown order, all hash tables are now correctly freed. - Improve behavior, when object system creation fails - Drop function NsfProfilePrint()