We open the log in ipa_webgui and this was being inherited by TurboGears
which uses the same log so everything was getting logged twice. Shut down
the log in ipa_webgui at the last possible moment. This will not catch
configuration errors.
Add a Not Found template.
Only print a traceback on 500 errors.
Don't read ipa.conf to get the realm, the kerberos libs do that for you.
Use the krbPrincipalName to change passwords
Make it possible to specify the principal at user creation.
Mail is not a required attribute so far, don't require it.
- Change sort functions to be on entities, so can use on the view pages too
- Fix bug: empty ajax search on useredit blows up
- Filter illegal characters from suggest uid/email methods
- Rename first/last name fields
- Make default font family sans-serif
- Speed up effect appear/fade rendering
- Add buttons to top and bottom of pages
- Make grouplist sortable
- Add noscript warning to welcome page
name and location of the keytab. In order for this keytab to be usable
TurboGears and Apache will need to run as the same user. We will also need
to listen only on localhost in TG.
kerberos principal name
Add an identity an visit class to TurboGears that can handle the user
without requiring a database
Update the UI to show the user correctly.
Note that this is currently disabled. It is hardcoded to always return the
principal test@FREEIPA.ORG in proxyprovider.py
It doesn't handle an unauthorized request because that can never happen.
Change account status to use select list and 'active'/'inactive' values.
Improve autosuggest to keep suggesting unless you overwrite a suggestion
(if you correct the name, it will re-suggest).
- Rename buttons
- Add fake "logged in as" text
- Increase font size and spacing for sidebar
- Fix search messages for no results
- Open ipa footer link in new window
This way it returns results even if the search times out.
The find_users() search now returns a counter as the first result, which
is set to -1 if the results are partial.
- "configurable" fields to search on
- tokenize search words
- prioritize exact matches over partial matches
- split match filter generation into a re-usable function.
Other updates:
- use finally block to return ldap connections
- update web gui to use new get_user methods
