The current LDAP provisioner will derive a gecos field with UTF-8 characters if the user's name contains accented characters (as an example). The resulting gecos field can't be successfully provisioned to a posixAccount object class that is based on rfc2307 (which allows ASCII characters only in the gecos field).

Backstory is there is that rfc2307 defines gecos with ASCII characters, and rfc2307bis (which was evidently never officially adopted), redefined this attribute to use UTF-8. Unfortunately there are many OpenLDAP installations that use rfc2307 and not 237bis. It would be helpful to have a checkbox for configuring the provisioner that would result in the gecos string being converted to to an ASCII approximation before being sent to the LDAP server. The equivalent python code for converting the string is something like: unicodedata.normalize('NFD', unicode_string).encode('ascii', 'ignore').decode('ascii')

Note that the voPosixAccount has no such problem since it also uses UTF-8 for the gecos field.