How Untested Software Affected Election Day

“Together we can restore the American dream.” That’s how Mitt Romney closed this video message directed at participants of Project Orca, a software project that aimed to light a fire under potential voters in order to get them to the polls on Tuesday. On Halloween, just days before the election, Romney released his message to volunteers detailing the special project that would give him an “unprecedented advantage on election day” using “state of the art technology.”

An ambitious undertaking, the smartphone application was intended to arm poll watchers with a way to record which voters were turning out to the polls by logging their names (okay, first they had to eavesdrop at voter check-in) and send the data back to a central server. Volunteers in Boston could then contact the potential swing state voters who hadn’t yet performed their civic duty. Not a bad concept, honestly. Kudos to the campaign for embracing mobile technology as part of a grass roots endeavor.

project orcaSo what went wrong? Well, pretty much everything. Volunteers were unable to test their PIN numbers or passwords until election day, which caused chaos when many PINs ended up being invalid. In the afternoon, as volunteers scrambled to fix the problem, reports started coming in from around the country that ORCA had crashed due to thousands of simultaneous log-ins. Add to this issues such as process gaps and communication gaffes – so even when the software worked, people didn’t know what they were supposed to do.

Photo source: Brian Cook Illustration

Frankly, Romney’s Project Orca ended up being his own personal #FailWhaleBut the election day software plague wasn’t isolated to the Romney camp. By mid-afternoon, stories and videos about voting machines that wouldn’t take Obama votes were already circulating the InterWebs. I was one of the lucky voters who lives in a small enough town that my voting booth is armed with a paper ballot and a black marker. I found that comforting after all the noise about the “software glitches.”

Oh wait, there’s more. Perhaps the most discouraging news stories came out of Ohio, a much-touted battleground state where the Ohio Secretary of State’s office updated their vote tabulation system with an apparently untested performance enhancement prior to election day. Of course, there was a resulting lawsuit that also pointed out the existence of a security hole in the form of a “back door” – a standard software device to allow developers or operations personnel to access a system without going through a standard authentication path. A judge ultimately threw out the lawsuit, but that isn’t even the point. The point is that anyone was cavalier enough to push untested software to a critical system.

Only a decade ago we were dealing with the fiasco that ensued with hanging chads in Florida during the 2000 Presidential election. And we thought that was bad. In today’s world, we’ve become so reliant on software to power our every gadget that we’ve stopped, in some senses, to take it as seriously as we should. We don’t just rely on code to provide us with games and productivity tools – we’ve incorporated it into the most powerful and important activities, all the way up to our presidential elections, where it’s a short skip from glitch to catastrophe.

In the end, it’s about restoring confidence – and no, I don’t mean in politics. I’m not going there, for which I’m sure you’re grateful. I’m talking about the confidence we feel in the systems around us and the software that powers them. It’s trite to point out that software is pervasive these days, so I won’t bother.

What is worth reflecting on is not how pervasive software is, but how powerful it is. Look at how much of our election day experience was powered by software applications, some of which were designed specifically for the election. But it does make you pause. When a country’s presidential election relies so heavily on software applications, there is a level of responsibility for making that software bullet-proof. Quality isn’t a choice anymore – it’s a requirement.

 



subscribe-to-our-blog





PRC-Free-Trial-Version-B



Comments

  1. Christian Jacob says:

    Well written und straight to the point. I am absolutely with you and further more: I noticed that the problem not only exist on the level of test managers not doing their job properly (if they even exist), but already fail at the level of a developer either “not having the time” for proper testing (meaning: the project manager simply didn’t consider a thing such as unit testing when planning the sprints) or (even more scaring) just doesn’t know hoe to unit test at all. It’s as if developers don’t learn how to write test code. But why is that so? Because they don’t want to? Because they don’t know that such a thing exist? Because they are “not allowed to” (I know how that sounds, but you know what I mean)… 
     
     
     
    I believe there is a lot of work to be done to get things back on the road…

  2. Ben Austin says:

    That’s a great point, Christian! As Lorinda points out, it seems a bit absurd to me that someone would push out untested software to any critical system. Thanks for the comment.

Speak Your Mind

*