When your code works but no idea how
@stux the test passes, but the product doesn’t because it relies on environmental conditions that aren’t necessarily documented.

@c0dec0dec0de @stux as a quality control guy, "it worked once, ship it" is not a standard I want to approve.

Me (mostly in my head): "Maybe if you took your torque training seriously, and didn't let your team break so many parts, you wouldn't be behind on your commits."

@Urban_Hermit @c0dec0dec0de @stux
Something like this happened at our company.
Tested one of our standard machines, yup pass. Once at the customer, the heater coil burns out immediately. We're in Europe, customer is in Australia, Murphy's law. Weeks later, coil arrives, service guy installs it. The thing burns out right away. Customer is royally pissed off, rightly so. They said we could have burnt down their facility with the faulty device.
(1/2)
@Urban_Hermit @c0dec0dec0de @stux
Turns out production accidentally reversed the connection for the heater coil, meaning it was under tension by default.
QC just tested: does it heat? Yup coil is hot. Does it cool? Yup valve opens.
We changed our protocols after that.
(2/2)