due to a mistake in the program Alaric wrote to test diode breakdown voltages, a test device ended up held in breakdown for ten seconds, with its current-limiting resistor only limiting it to 50 watts, way too much for a tiny TO-220
Somehow, despite the fact that the tin plating melted and the heatsink mounting tab now has a mirror finish, indicating that it must have gotten to at least a temperature of 232°C, with the junction temperature even higher, it seems to still work perfectly fine
-F