1778[BUGFIX] Removed dependency on TYPO3_MODE in saltedpasswords
Commit message
Review
-
Code review CR +2
- Dmitry Dulepov +2
- Stefan Neufeind +1
- Markus Klein +1
- Steffen Ritter +1
-
Verified V +2
- Dmitry Dulepov +2
Issues 1
Files 1
Review log 22
-
Björn Pedersen commented on patch set 2 10/06/11, 07:37
Hm, why do you need to change that? -
Björn Pedersen commented on patch set 2 10/06/11, 07:37
I would prefer that you didn't submit this Hmm, that is something which the current behaviour does not allow. And to my opinion correctly. The security team should give a statement here. -
Björn Pedersen commented on patch set 2 10/06/11, 07:37
Well, he suggested moving the tet to initAuth, where the parent object is available. -
Björn Pedersen commented on patch set 2 10/06/11, 07:37
No score -
Dmitry Dulepov commented on patch set 2 V 0 10/06/11, 07:37
by team member; Looks good to me, approved Tested and works ok. -
Stefan Neufeind commented on patch set 2 10/06/11, 07:37
Looks good to me, but someone else must approve -
Stefan Neufeind commented on patch set 2 10/06/11, 07:37
Marcus? Björn? -
Markus Klein commented on patch set 2 10/06/11, 07:37
Looks good to me, but someone else must approve Since Marcus Krause is a member of the security team, you already have an opinion - he suggested it. -
Steffen Ritter commented on patch set 2 10/06/11, 07:37
Looks good to me, but someone else must approve -
Marcus Krause commented on patch set 2 10/06/11, 07:37
No score You said you would update the patch. I'll wait for the new version. -
Marcus Krause commented on patch set 2 10/06/11, 07:37
Verified; Looks good to me, but someone else must approve I'm fine with this change! -
Marcus Krause commented on patch set 2 10/06/11, 07:37
No score -
Alexander Stehlik commented on patch set 2 10/06/11, 07:37
It's Marcus, not Markus, sorry for that! -
Alexander Stehlik commented on patch set 2 10/06/11, 07:37
Hi Markus, thanks for looking into this. I would really like to hear your or Björn's opinion which solution you would prefer after I posted my reasons why I used the init() method instead of the initAuth(). If you think initAuth() is the way to go I will update the patch. Otherwise you can vote for the current version. -
Alexander Stehlik commented on patch set 2 10/06/11, 07:37
You are right, Markus suggested to use the initAuth method. There are 2 reasons I solved it differently: First, it seemed cleaner, to leave the detection, if the service is available, in the init method, since this is the expected return value of this method. Second, I didn't want to touch the authentication logic. But now, after thinking of it, it should be quite straightforward. Move the detection isUsageEnabled detection in the initAuth method and store it in a class property. In the authUser method I check at the very beginning, if isUsageEnabled is TRUE, otherwise I will return 100. If you think this is better I will update the patch. -
Alexander Stehlik uploaded patch set 2 10/06/11, 07:37
-
Alexander Stehlik commented on patch set 2 10/06/11, 07:37
Just CGL cleanup. Sorry Mr. Jenkins ;) -
Alexander Stehlik commented on patch set 2 10/06/11, 07:37
The TYPO3_MODE constant is not always correct. I have a scenario where a kind of single sign on extension for FE/BE users (similar to the simulatebe Extension) is used. The user logs in at the frontend and the login data is forwarded to additionally authenticate the user for the backend. The TYPO3_MODE constant is always "FE" but the user that should be authenticated is a backend user. -
Alexander Stehlik commented on patch set 2 10/06/11, 07:37
Thanks for your feedback Björn. You are right, this should be reviewed by a security team member since it affects the login mechanism. At the moment, forwarding login data is already possible (with or without salted passwords). It works without any problems. So this isn't a change this patch introduces. The only problem is, that passwords, that have not been salted yet will not be updated in the correct table. Since the parent t3lib_userAuth object uses it's $user_table property for determing from which table the user information should be read, I don't see a reason why the saltedpasswords extension shouldn't do the same. This is actually more consistent. The other part of the patch affects the mechanism, that detects if saltedpasswords are enabled for the current login (BE or FE). I introduced this based on Marcus Krauses comment in the issue tracker. -
Mr. Jenkins commented on patch set 1 10/06/11, 07:37
Fails; I would prefer that you didn't submit this Build Failed http://ci.typo3.org/job/typo3-v4-core-gerrit/377/ : FAILURE -
Mr. Jenkins commented on patch set 2 10/06/11, 07:37
Build Successful http://ci.typo3.org/job/typo3-v4-core-gerrit/378/ : SUCCESS -
Dmitry Dulepov merged patch set 2 10/06/11, 07:37