Comment 13 for bug 1722185

Since the proposed changes are quite big and seem a bit intrusive, I would appreciate a Regression Potential analysis as per our SRU guidelines. I would like to know where we could expect regressions after this change lands in the stable series.

Also, even though this does seem to be like a frequently-seen crash for both bionic and artful, I wouldn't want to jam it into the stable series too hastily. I see the second PackageKit pull request has not yet been approved by upstream - could we get someone reviewing it and merging before we proceed? Normally I'd also wait for cosmic to migrate, but I actually see we're only blocked on the armhf tests not being run - so it seems to be fine so far.