[Tickets #10433] Re: Successful password change leads to infinite loading of site

[bugs at horde.org]

DO NOT REPLY TO THIS MESSAGE. THIS EMAIL ADDRESS IS NOT MONITORED.

Ticket URL: http://bugs.horde.org/ticket/10433
------------------------------------------------------------------------------
  Ticket             | 10433
  Updated By         | stephan at admin.nabira.de
  Summary            | Successful password change leads to infinite loading
                     | of site
  Queue              | Passwd
  Version            | Git master
  Type               | Bug
  State              | Duplicate
  Priority           | 1. Low
  Milestone          |
  Patch              |
  Owners             |
+New Attachment     | Passwd.php.diff
------------------------------------------------------------------------------


stephan at admin.nabira.de (2011-08-17 11:59) wrote:

> Not to offend you in any way but the diff file is empty.
Whatever happened to the file I do not know, but I will try another upload.

> You should have mentioned it is about imp application auth. Other  
> backends are working without problems so unconditional logout is no  
> good option for these cases.
Ehm, no. It is about Horde auth using IMAP driver. I am not using imp  
directly as authenticator,  although Horde may do that internally,  
which I do not know.

My config regarding this is:

horde4/config/conf.php:
...
$conf['auth']['params']['hostspec'] = 'mail.hostname.de';
$conf['auth']['params']['port'] = 143;
$conf['auth']['params']['secure'] = 'tls';
$conf['auth']['driver'] = 'imap';
...

horde4/imp/config/backends.php:
...
$servers['imap'] = array(
     'disabled' => false,
     'name' => 'IMAP Server',
     'hostspec' => 'mail.hostname.de',
     'hordeauth' => false,
     'protocol' => 'imap',
     'port' => 143,
     'secure' => 'tls',
     'maildomain' => '',
     'cache' => false,
);

>
> What passwd backend is used here?
MySQL. I write to the same DB the IMAP-Server uses as authentication  
database. I have tried authenticating against the DB directly but  
regarding this problem there is no change in behaviour.