Review main Merged

26697[BUGFIX] Functional tests cannot write to backend log

Oliver Hader CR +2 V +2 Easy +1 −1 updated 01/23/14, 13:48

Commit message

Functional test cases cannot write to the backend log. The
log entry also has an information about the currently used
IP address, which is taken from the environment variable
REMOTE_ADDR. In a CLI disposal this value is NULL and cannot
be stored in the database, since the accordant sys_log.IP
field does not support NULL values.

Review

  • Code review CR +2
    • Wouter Wolters +2
    • Anja Leichsenring +1
  • Verified V +2
    • Wouter Wolters +2
    • Anja Leichsenring +1

Files 1

Review log 9

  1. Anja Leichsenring voted on patch set 1 V +1 CR +1 01/08/14, 19:29

  2. Wouter Wolters voted on patch set 1 V +2 CR +2 01/08/14, 19:35

  3. Wouter Wolters merged patch set 2 01/08/14, 19:35

  4. Mr. Jenkins commented on patch set 1 01/08/14, 19:42

  5. Michael Stucki commented on patch set 2 01/08/14, 20:13

    That was a bit too fast IMHO. I was discussing the change with Oliver while you merged it :) The change is just a workaround, where the real problem is not fixed. My suggestion is to return 127.0.0.1 in getIndpEnv() if CLI_MODE is set. That would also fix other potential problems with the missing IP.
  6. Oliver Hader commented on patch set 2 01/08/14, 20:17

    The patch is not wrong since the field type to be stored shall be string and not NULL... Michael, you can go ahead and create a new issue for defining server environment variables if called in CLI context.
  7. Wouter Wolters commented on patch set 2 01/08/14, 20:35

    How could I know that you were dicussing this, I was on IRC and Skype?? But IMO this was valid and therefor I merged it. Your suggestion looks more a FEATURE. This solves the problem in a way we can't argue about. Other default values for this is another story
  8. Michael Stucki commented on patch set 2 01/08/14, 20:47

    Well Wouter, you have a point here :-) I was in a direct conversation with Olly, you couldn't know that... I agree that it could be considered as a feature. For now I have reported a new issue for it at least: http://forge.typo3.org/issues/54850
  9. Oliver Hader commented on patch set 2 01/23/14, 13:48

    I'm back-porting this to other branches since the breaking change with NULL values has been back-ported to older branches as well...