https://bugs.winehq.org/show_bug.cgi?id=25229
--- Comment #19 from Rosanne DiMesio dimesio@earthlink.net --- (In reply to Jens Reyer from comment #16)
This just hit me again, so this bug is probably not fixed.
I had rejected https://appdb.winehq.org/objectManager. php?sClass=version&iId=26175&iTestingId=103092
I found it in the rejected queue, but when I tried to accept it from there, the AppDB threw up an error message that said "Database Error: Table 'apidb.distribution' doesn't exist." The test report disappeared from the rejected queue but turned up back in the submitted queue, and I accepted it from there. So there's definitely a bug in processing reports from the rejected queue, but that's separate from what's reported here.
As for this bug, what I think has been happening is that some users have edited their reports but failed to resubmit them, so they were never moved out of the rejected queue, but maintainers have been interpreting the email notice about modifying the test to mean that it was resubmitted. If a test report was actually resubmitted, and not just edited, then the notice would say it was submitted. At least, that's what I found in my local test environment, and that's been my experience as a maintainer.
We probably shouldn't be sending notices about modifications until the report actually is resubmitted.