Confirmed the behavior, and indeed with bug LP: #2038651 fixed, it actually produced *triple* the results it should.
Problem was it was double-looping over the results. Thanks for reporting this, I'm surprised it didn't get caught earlier.
Confirmed the behavior, and indeed with bug LP: #2038651 fixed, it actually produced *triple* the results it should.
Problem was it was double-looping over the results. Thanks for reporting this, I'm surprised it didn't get caught earlier.