> Hm... In my case, this most likely going to all be done for anonymous > binds > (Outlook email client comes to mind). My though here is to create a fake > branch in the DIT (cn=outlook,dc=stanford,dc=edu) that rewrites Stanford's > custom schema into what Outlook (or another email client) wants.
Then it would be (almost) trivial: use back-ldap (2.2) or back-relay with slapo-rwm (2.3) to map objectClasses/attributeTypes the way Outlook wants them. If you use back-ldap on a different host you may keep the real naming context; if you use back-ldap on the same host, or back-relay you'll need to rewrite the naming context as well. All you'd likely need to do is: <2.2> database ldap suffix <virtual naming context> uri ldap://<host:port stuff> suffixmassage <virtual naming context> <real naming context> map attribute <outlook> <real> ... map objectclass <outlook> <real> ... </2.2> <2.3> database [relay|ldap] suffix <virtual naming context> # if back-ldap... uri ldap://<host:port stuff> # if back-relay relay <real naming context> overlay rwm rwm-suffixmassage <real naming context> rwm-map attribute <outlook> <real> ... rwm-map objectclass <outlook> <real> ... </2.3> > I myself actually don't need the uid/userID thing anymore (someone else @ > Stanford had asked the question, but it never became an issue). fine. p. -- Pierangelo Masarati mailto:[EMAIL PROTECTED] SysNet - via Dossi,8 27100 Pavia Tel: +390382573859 Fax: +390382476497
