Fri, 10 Apr 2026 14:48:39 +0000 https://bugs.horde.org/ticket/3390 ldap code overwrite last_login pref I use ldap backend for preferences. In horde+imp last_login works. In horde+imp+ingo last_login does not work anymore. When imp loads preferences, it load horde preferences as well. In that step it set last_login with the ldap value, "display" last_login value and set last_login to current time. When ingo loads preferences, it load horde preferences as well. In that step it set last_login with the ldap value and the last_login value from imp is lost. If my workaround is usefull you can use it, modify it, ... ------------------------------------------------------------ *** ldap.php.orig Thu Feb 2 09:41:30 2006 --- ldap.php Thu Feb 2 14:34:01 2006 *************** *** 382,389 **** /* Retrieve this preference. */ if (isset($this->_prefs[$pref])) { ! $this->_setValue($pref, base64_decode($val), false); ! $this->setDefault($pref, false); } else { $this->add($pref, base64_decode($val), _PREF_SHARED); } --- 382,398 ---- /* Retrieve this preference. */ if (isset($this->_prefs[$pref])) { ! # when imp loads, the last_login is read from ldap ! # then imp set new value for last_login, but it does not write it to ldap yet ! # then ingo loads and last_login is read from ldap: last_login is overwriten with old value ! # so we need to check if last_login is globaly dirty and use global last_login value instead ! if ($GLOBALS['prefs']->isDirty($pref)) { ! $this->_setValue($pref, $GLOBALS['prefs']->getValue($pref), false); ! $this->setDefault($pref, false); ! } else { ! $this->_setValue($pref, base64_decode($val), false); ! $this->setDefault($pref, false); ! } } else { $this->add($pref, base64_decode($val), _PREF_SHARED); } Thu, 02 Feb 2006 14:05:54 +0000 https://bugs.horde.org/ticket/3390#t16162 I think this is actually a general flaw in the Prefs code, in that it will load Horde preferences regardless of whether they're already loaded. This is a waste and also, as you diagnosed, can result in overwriting unwritten values. I'm still puzzling out the best fix; your way does work, but I think we'll arrive at something a bit different eventually. Needs to be fixed for at least the SQL driver as well, possibly more. Thu, 02 Feb 2006 19:55:32 +0000 https://bugs.horde.org/ticket/3390#t16186 This is exacly the same bug I''ve metioned in bug #2838 und #2829 Both marked as "bogus" !?!?! Because some "Horde-people" (e.g. Michael Slusarz) are not able do find out differences between bufg #2838 und #2829 the second bug is state bugus. However bug #3390 is marked as bug with priority HIGH although a real simple bugfix (and the final solution for this problem) was provided in bug #2838 (163 days ago)! Nevertheless this would be my last post on this stupid buf tracker list. LecksMichAmOrsch Thu, 06 Apr 2006 20:42:36 +0000 https://bugs.horde.org/ticket/3390#t18630 Duplicate of bug 2838. Sun, 09 Apr 2006 00:11:33 +0000 https://bugs.horde.org/ticket/3390#t18698