Thanks a lot Bjørn and Alan. >> I believe it is time to revisit the Ascend dictionary fixup. > > I think so, yes. > >> Yes, I can see the point that this will break existing setups. But I >> will argue that such setups have *always* been broken. At least when we >> are discussing the FreeRADIUS 2.x era. You cannot support them without >> breaking RFC conformance, which should have a higher priority even in >> the stable tree. > > Pretty much, yes. > >> Or just rename the next 2.1.x release "2.2.0" if that makes you feel >> better :-) > > No... 2.2.0 is a different branch in git, and has major new features, >like TCP transport, and many more which are waiting for certain things >to happen. > > For 2.1.11, the safer alternative is to move the "illegal" attributes >to their own dictionaries, and then include those *before* the RFC >dictionaries. People will still be able to create/send attributes using >the illegal names. But when attributes are received, the new names will >be used by default. > > That *is* a change in behavior. But it's time for 10-year-old >equipment to be deprecated.
I confirmed that this problem was solved by not "Don't include Ascend dictionary" but "Include Ascend dictionary before RFC dictionary". Thanks for your help. ichiro tanaka - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html