#106743Creating new BE users now requires additional authentication
Description
The change introduced with Require step-up authentication for password change and https://forge.typo3.org/issues/103252 requires authentication when changing password.
IMHO, this is unnecessary when creating new users and can be quite a nuisance for setting up new users.
Side note:
(Additionally, for most users we don’t even need the password because we use external login server (SSO). We also have introduced a field where you can select the login type. It would probably make sense to remove the password field, if native login is not supported, however I would prefer if a mechanism for this was introduced in core, since extensions implementing this would probably result in conflicts if you want to provide several.)
Versions
- change was applied to >= v12
- noticed change in v12
Reviews 12
- 12.4 91860[DOCS] Remove unused class import in code example
- 13.4 91859[DOCS] Remove unused class import in code example
- main 91842[DOCS] Remove unused class import in code example
- main 89642[DOCS] Fix not displaying Changelog entry for 106743-SudoMode
- 12.4 89684[DOCS] Fix not displaying Changelog entry for 106743-SudoMode
- 13.4 89683[DOCS] Fix not displaying Changelog entry for 106743-SudoMode
- 12.4 89616[BUGFIX] Fix indefinite sudo confirmation when copying backend users
- 13.4 89615[BUGFIX] Fix indefinite sudo confirmation when copying backend users
- main 89597[BUGFIX] Fix indefinite sudo confirmation when copying backend users
- 12.4 89564[FEATURE] Introduce Sudo-Mode Events
- 13.4 89563[FEATURE] Introduce Sudo-Mode Events
- main 89491[FEATURE] Introduce Sudo-Mode Events
Related issues 2
Files 1
History 919
-
Sybille Peters commented <span data-processor="relativetime" data-value="2025-05-22T15:46:02+00:00" > 05/22/25, 15:46 </span>
Oliver:
> Would that be a feasible alternative for your use-case?Yes, theoretically yes (did not test the patch yet). Thank you for responding.
-
Gerrit Code Review pushed patch set 3 for main to change 89491 status
NewUnder Review <span data-processor="relativetime" data-value="2025-05-22T18:11:19+00:00" > 05/22/25, 18:11 </span> -
Gerrit Code Review pushed patch set 4 for main to change 89491 <span data-processor="relativetime" data-value="2025-05-25T07:30:59+00:00" > 05/25/25, 07:30 </span>
-
Gerrit Code Review pushed patch set 5 for main to change 89491 <span data-processor="relativetime" data-value="2025-05-26T09:17:05+00:00" > 05/26/25, 09:17 </span>
-
Oliver Hader commented <span data-processor="relativetime" data-value="2025-05-26T09:44:33+00:00" > 05/26/25, 09:44 </span>
I wanted to leave feedback on the assumption in the original issue description:
Sybille Peters wrote:
> IMHO, this is unnecessary when creating new users and can be quite a nuisance for setting up new users.The basic concepts of the underlying problem are referenced in https://cwe.mitre.org/data/definitions/620.html
Allowing to create new users without verification would allow an attacker to create a new admin user which can be used further.
This also applies when creating new non-admin users, since the it would allow to grant privileges generously. -
Gerrit Code Review pushed patch set 6 for main to change 89491 <span data-processor="relativetime" data-value="2025-05-26T12:51:43+00:00" > 05/26/25, 12:51 </span>
-
Gerrit Code Review pushed patch set 7 for main to change 89491 <span data-processor="relativetime" data-value="2025-05-26T13:26:57+00:00" > 05/26/25, 13:26 </span>
-
Gerrit Code Review pushed patch set 8 for main to change 89491 <span data-processor="relativetime" data-value="2025-05-26T19:20:02+00:00" > 05/26/25, 19:20 </span>
-
Oliver Hader changed parent task #106772 <span data-processor="relativetime" data-value="2025-05-26T19:26:59+00:00" > 05/26/25, 19:26 </span>
-
Gerrit Code Review pushed patch set 9 for main to change 89491 <span data-processor="relativetime" data-value="2025-05-27T11:48:19+00:00" > 05/27/25, 11:48 </span>
-
Gerrit Code Review pushed patch set 1 for 13.4 to change 89563 <span data-processor="relativetime" data-value="2025-05-27T12:11:29+00:00" > 05/27/25, 12:11 </span>
-
Gerrit Code Review pushed patch set 2 for 13.4 to change 89563 <span data-processor="relativetime" data-value="2025-05-27T12:15:57+00:00" > 05/27/25, 12:15 </span>
-
Gerrit Code Review pushed patch set 1 for 12.4 to change 89564 <span data-processor="relativetime" data-value="2025-05-27T12:17:52+00:00" > 05/27/25, 12:17 </span>
-
Oliver Hader commented and changed status
Under ReviewResolved done0 %100 % <span data-processor="relativetime" data-value="2025-05-27T13:25:07+00:00" > 05/27/25, 13:25 </span>Applied in changeset commit:76d48d96e646111ea6681abcc19225410edca96b.
-
Garvin Hicking changed Related to #106795 <span data-processor="relativetime" data-value="2025-05-29T18:41:30+00:00" > 05/29/25, 18:41 </span>
-
Sybille Peters commented <span data-processor="relativetime" data-value="2025-06-02T07:39:24+00:00" > 06/02/25, 07:39 </span>
Feedback after patch: I debugged the new version and it looks like in DataHandlerAuthenticationContext::processDatamap_postProcessFieldArray that it would be able to detect that a new BE user is being created (dataHandler->datamap->be_users[NEW….) if BE user is created via normal process. In this case, I don’t really see why it should be necessary to query the sudo login if a new user is being created and a new password assigned.
I understand we can now use the events but it makes adding code necessary.
(sorry, I should have commented this before, but I am currently still a bit disabled after an accident).
-
Elias Häußler changed Related to #108283 <span data-processor="relativetime" data-value="2025-11-24T07:19:38+00:00" > 11/24/25, 07:19 </span>
-
Benni Mack changed status
ResolvedClosed <span data-processor="relativetime" data-value="2025-12-02T13:59:11+00:00" > 12/02/25, 13:59 </span>
https://review.typo3.org/c/Packages/TYPO3.CMS/+/89491 would introduce a new event that allows to by-pass the step-up authentication. Would that be a feasible alternative for your use-case?