In reply to a post by Zarjaz:

That's because there was one.

I'm afraid to say that's not always the case. CP's will test and test and test again until they manage to record the one test that 'comes back with an odd result',

Is that not a fault then? Surely they are using Openreach's tools?