The culture among software developers today makes quality secondary to delivering the product on time. They (usually) make a good faith effort to produce good work, but don’t kid yourself: a lot of the people creating software aren’t that smart, and don’t have particularly high standards for their own work. They are used to “letting the customer be the tester”.
This is reason enough to absolutely minimize the use of computer tools in elections, until very high standards are met and proved through exhaustive testing.
This isn’t true only of software today. “Vaccines” developed in haste are forced on the public under the rules of “emergencies”, and in many cases no one knows what the broad effects may be.
In medicine too, we have to recognize that imagined “emergencies” are not sufficient justification to bypass exhaustive testing. And if it must be bypassed, it should be done with ENORMOUS warnings and effort to restrict use to the only the most dire cases that constitute a real emergency. And once the emergency has passed, emergency authorizations should be withdrawn too.
But in some ways, the threat posed by poor election quality is greater than that posed by a pandemic. It enables mistakes in the best case, and cheating in the worst.
No useful protocols exist for dealing with flawed elections.
Lock, load, fire, repeat as necessary is a fine protocol to address fraud. Corrupt election officials have homes somewhere in Arizona