Our Openfire is set up to use the LDAP authentication provider. We use Miranda as the client (but the Spark behaviour is identical).
The problem occures whenever the password contains the extended ASCII characters, such as § or ü, etc.
The LDAP server expects the UTF-8 encoded string for the password but gets the ISO-8859-1 characters instead and naturally does not work.
If we convert the ISO-8859-1 string into UTF-8 and paste the resulting bytes into the password field of the client then the authorization is successful.
The question is can Openfire take care of this conversion? I believe, LDAP uses UTF-8 generally, so the LDAP connection provider should take care of that, shouldn’t it?
So… nobody can give any comment on this issue?
I cannot beleive we are the only ones suffering from the problem above. Are there any other people out there using Openfire with LDAP authentication back end?
It actually got worse with Openfire 3.5.x.- the workaround that worked before:
If we convert the ISO-8859-1 string into UTF-8 and paste the resulting
bytes into the password field of the client then the authorization is
… is not working anymore. The users who have extended ASCII characters in their passwords cannot use Openfire/Jabber at all.
Can anybody from the developers please give any feedback? We are going to roll back to pre 3.5.x version for now but in the long run we would need to consider another solution.