Yes, of course I agree; and I would hope that PG is aware of design techniques that help to minimise the likelihood of introducing errors. But none the less, errors can still happen; testing inherently cannot be complete and perfect. Any particular error may only happen in rather specific circumstances - as another example, consider the early Pentium processors that were manufactured with a fault in the floating-point divide instruction: this was a blatant and readily demonstrable fault, yet one that even the resources of Intel managed to miss in testing. In any case, given an actual fault for whatever reason, a sufficiently measured response should be made, to minimise the possibility of some other error being made by rushing the job.planetgroove wrote:But rendering audio material with dropouts is one of the most severe faults, an audio editor can have. If everything else goes wrong - that at least should work 100% flawlessly.
I think you would agree, right?