Comment 8 for bug 1738412

On Thu, Mar 15, 2018 at 11:42:49PM -0000, Tommy Nevtelen wrote:
> On 2018-01-02 15:38, Robie Basak wrote:
> >> This is a very serious issue...
> > Why is this "a very serious issue"? It seems to me that it only affects
> > invalid configurations so should only affect development and testing of
> > deployments rather than in production? The workaround (if it can be
> > called that) is to fix the invalid configuration.
> I kind of hope that this is a joke :)

Sorry, no, it wasn't a joke.

> You know sometimes things don't go as intended, that's why the check is
> there.

I don't disagree. It is a bug and as I said I have no objection to
getting it fixed in older stable releases if somebody wants to drive
that. But I maintain it is not a serious issue, because as I explained
it only causes some inconvenience to users during development and
testing and does not affect production systems. Claiming it is a serious
issue dilutes attention from issues that really are serious, which is
why I objected to that statement in this case.

> So it has been a while now. What's the next step to get this in?
> The fix is really trivial. Can I help with something to get it rolling?

Please see The
fix may be trivial, but unfortunately updating the stable release is
not. The risk of regressing users who expect stability means that we
have to take considerable care. For example, an unrelated latent bug
could be exposed by a rebuild. See for details.