Alex Karasulu a écrit :
Emmanuel is correct here since JNDI is far from being consistent about LDAP.
Keep in mind that RDN *should* have been exactly what Alex said it was, ie a *relative* (to a context) name, so 'x=a, y=b' sounds like a natural candidate for a RDN. This is puzzling that it's not the case ... It would have been better to use ATAVs for RDN, RDN for what it seems to be, RootContext for the, hum, root context, and DN for the whole thing. Ok, enough insanity ;)
I would like to thanks all the people who successfully drove us mad with all those names ;)
Emmanuel
