On Mon, Feb 18, 2002 at 05:08:52PM +0100, Prune wrote:
Chris Lightfoot wrote:
On Mon, Feb 18, 2002 at 03:44:11PM +0100, Prune wrote:
What it does :
-do auth agains an ldap server
-get the location of the mailbox (or maildir) from LDAP
-get the uid/gid of the mailbox from LDAP
OK, this all looks sensible. I take it that the way that
authentication is done is defined by LDAP, so that you
don't have to retrieve a password from the directory
explicitly?
right. That's why it's a good thing to use TLS, so data from the client
to LDAP are encrypted overt the network.
Ldap has a special way to authenticate users with a methode called 'bind'.
First you connect to the server.
Then you 'bind' as manager (privilegied read user).
you search for the user and his attributes
once you have all this, you can 'bind' again as the user.
I'm curious why someone would require a privileged user to perform
the mail -> uid/DN search? In other words, what would be the
advantages of putting access controls on a mail attribute? It seems to
me to defeat one of the original purposes of LDAP, e.g. address books.
(Perhaps I'm missing something here).
I would have expected this to be obtainable from an anonymous
bind/search which is quicker than an authenticated bind.
the main problem is that anybody can access to you user listing / mail.