I agree with you, from the point of view of the rpm API a simple integer value might not be sufficient for encoding the status a massive rpm operation.
However, for a batch script, a simple warning message is not sufficient to tell that your software upgrade went wrong due to a failure in the %post scriptlet
leaving you mission critical application stopped due to simple typo in a configuration file...
So I think rpm (yum behaves the same way) should treat errors in %post script as before or, at least, it should be possible to choose the old or the new approach with a macro or via a command line switch.
I agree with you, from the point of view of the rpm API a simple integer value might not be sufficient for encoding the status a massive rpm operation.
However, for a batch script, a simple warning message is not sufficient to tell that your software upgrade went wrong due to a failure in the %post scriptlet
leaving you mission critical application stopped due to simple typo in a configuration file...
So I think rpm (yum behaves the same way) should treat errors in %post script as before or, at least, it should be possible to choose the old or the new approach with a macro or via a command line switch.