----------------
- In order to avoid duplicate search results, the enumerateUsers
method used a simple dictionary to store DNs for records that
were already processed. However, the keys put into this dictionary
were munged and really could not be compared to raw search
result DNs anymore. Thanks go to Wichert Akkerman for spotting this
obvious error (http://www.dataflake.org/tracker/issue_00485).
- Speed up enumerateGroups by letting the LDAP server do more of
the filtering (thanks to Wichert Akkerman,
http://www.dataflake.org/tracker/issue_00483)
- Applied a performance fix to the ActiveDirectoryPlugin's
_recurseGroups method (thanks got to Mark Hammond for the patch,
http://www.dataflake.org/tracker/issue_00476)