Opened 5 years ago

Kerberos support

Reported by: Owned by: raubvogel@… airadier normal TracWikiPrintPlugin normal 0.12

Description

Are you planning on supporting Kerberos for authenticating the server itself? Or perhaps using httpauth_user and httpauth_password to send, well, the username and password for the user to kerberos (since mod_auth_kerb will accept them instead of just tickets)?

comment:1 Changed 5 years ago by airadier

In the Wikiprint options page in Admin you can set a username and password to use in httpauth_user and httpauth_password. Does it work for you or is there anything else needed in order to make it work?

comment:2 follow-up: ↓ 3 Changed 5 years ago by raubvogel@…

AFAIK, one of my problems is that I am accessing trac using mod_alt_kerb, which does not seem to failover nicely to another auth system like basic auth (it will but then it will not accept tickets). So I was thinking on just having everyone use kerberos and be over with it.

To test that, I am editing wikiprint.py to something like this (if it works then it is easy to edit/expand the if statement to handle kerberos in addition to simple auth):

            #Allow wikiprint to authenticate using user and password, Basic HTTP Auth or Digest
if self.env.config.get('wikiprint', 'httpauth_user'):
self.env.config.get('wikiprint', 'httpauth_user'),
auth_handler = urllib2.HTTPBasicAuthHandler(pwmgr)
auth_handler2 = urllib2.HTTPDigestAuthHandler(pwmgr)

opener = urllib2.build_opener(auth_handler, auth_handler2)
urllib2.install_opener(opener)


But I (probably) also need to add the urllib_kerberos stuff before testing.

comment:3 in reply to: ↑ 2 Changed 5 years ago by raubvogel@…

Just to let you know, so far I have not been able to have tracwikiprint to authenticate against kerberos with an unchanged wikiprint.py and using only httpauth_user and httpauth_password. This is why I am trying other avenues.

comment:4 Changed 5 years ago by airadier

• Status changed from new to assigned

Related to #8622