Comment 7 for bug 2040059

A no-change rebuild would be perfectly acceptable, but the more-or-less arbitrary bump in the text-failure limit makes me nervous. On the other hand, it's already "200" by default (!) so one might argue: if we're already ignoring 200 tests arbitrarily, what's 30 more non-specific failures? (of course, one might also argue: if we blindly ignore 200 non-specific tests ... why bother running them?).

However, the regression risk seems reasonable given that the current state is that the application won't even launch, and we obviously don't want to break things on upgrade. So, I'll do a full test-build locally (to get a feel for what sort of "expected failures" occur), and sponsor.