In message <1357163025.25.0.816891681034.issue2550731@psf.upfronthosting.co.za>
 <1357163025.25.0.816891681034.issue2550731@psf.upfronthosting.co.za>,
Bernhard Reiter writes:
>Bernhard Reiter added the comment:
>I've poked around in the code and I'm attaching a hg export
>  issue2550731-20130102ber.patch
>it is NOT TESTED (as I ran out of time)
>a) I've changed the states and the variable name
>b) From browsing the code it seems that the email-sign-openpgp place
>   may not be working correctly in all cases. Emails could be only
>   encrypted or encrypted and signed. I'm not sure of reaching this code
>   point always means they have been sucessfully verified.
>c) hg export applies to 4717:9dc50be521ee
>
>Was the fact that you were using a remote postresgl database
>important for the reproduction of your unwanted behaviour?

Yes it was.

When I get a chance to look at the patch, I'll comment back here.

One idea I also had was to create a testing tracker with appropriate
fields (e.g. arrivalMethod fields in the msgs and issue) that I could
change from the auditor. This would be required for unit testing
purposes anyway.