>> And: Non-enterprise users will most likely not need AD/LDAP support. > I strongly disagree with this. Everyone with more than 2 users in their organization

>> And: Non-enterprise users will most likely not need AD/LDAP support. > I strongly disagree with this. Everyone with more than 2 users in their organization can benefit from AD/LDAP support, if it is easy to set up and administer. Actually anyone with either more than one user, or more than one device/computer IMNHO. We really (still) need a sane, easy-to-set up authz/auth solution. Things have gotten a lot better - but still, nfsv4 is sort of not here, samba/cifs isn't quite secure, coda is sort abandoned, davfs isn't quite fully-featured... And so on and on. Ssh keys are (almost) easy to use, but really hard to manage securely (everyone just ignores this, and pretend private keys won't leak) - managing a ssh CA is essentially as tricky as any other CA. Ms realised this, and boxing up kerberos+ldap+dns (and some other bits) was their solution: Active Directory. I'm not convinced we can't do better for a turn-key linux-first, modern solution (we have public key cryptography now, that should(?) simplify a kerberos work-a-like? Make it a standard, and bsd support should be easy, and a port to Windows/gina feasible. I strongly believe the particulars of the solution doesn't really matter much - just make it open, with a good test suite and decent reference implementation. Redhat is AFAIK doing great work here, as was/is Debian Edu (nee "skolelinux").