The Third Bit

To nitpick, in this case it might not have necessarily been a software glitch:

“When the call was disconnected — for undetermined reasons — the call centre contacted emergency services based on the southern Ontario address they had in the system for the family.”

http://www.canada.com/calgaryherald/news/city/story.html?id=7353b0e4-5654-442c-b28a-54d1e7c1cbe1

The phone disconnection could have been the family hanging up too early.

By Andrew Louis on May 2, 2008

They are certainly responsible on some level, but being held accountable? That seems absurd. There is a whole chain of people that are all involved, starting with the person who introduced the bug (assuming it is a bug) and ending with the people at the call centre who did not get an ambulance from the town the call was in.

I don’t understand why the 911 call center did not react to this, and if anybody should be held accountable, it’s certainly them.

By Lennart Regebro on May 2, 2008

The general rule is that management is 80-90% responsible for all decisions in the software industry, therefore they are responsible for the outcome. That includes the rush job management forced on the programmers who had to write the system that connected to the 911 facility. That rush was due to poor planning by management who didn’t think they’d need to support 911 and most likely considered it a cost.

Additionally, I bet if you dug further you’d find that management also controlled pay, qualifications, training, quality, testing procedures, technology used, language choices, documentation requirements, tools used, work environment, temperature, and who should work on what when and how.

If we want to start making programmers responsible, then management needs to give them full control of the process so that they can be made responsible. Since management is actually in control, they are responsible.

By Zed A. Shaw on May 3, 2008

Not that I’m really surprised by the debate, but no, the developers are not responsible.

Having around VOIP for over 10 years now from a watch-the-tech-evolve, 911 service has always been an issue. The system behaved exactly as it should have and gave them a 911 operator who dispaced an ambulance to the address of record. VOIP is by design geographically neutral. To go (admittedly) extreme, if anyone is at fault it is the parents for
a) not understanding what they signed up for
b) not changing their address with a service provider

Note the faq about 911 on roger’s site: Where the service is available, Rogers Home Phone comes complete with enhanced 911 service. Enhanced 911 means that your street address will be automatically delivered to the emergency operator if you ever have to place a 911 call.

Remember back in the day when geographic location for cell phones didn’t exist for 911? When asked whether people would give up their land lines for a cell phone the single people were ’sure, bell sucks’ but the parents would say ‘no, it is not safe for me to not have 911 service’. Well, guess what, VOIP is the same thing, only attractivly priced.

Is it a tragedy? Of monsterous proportions. Could it have been prevented by another day of testing? More unit tests? Management pushing the release date out? No. Could it have been prevented by the parents updating their address with rogers? Maybe.

Here is the link to the explanation of vonage’s 911 offering. Seems to me they had a couple similar situations too http://www.vonage.com/features.php?feature=911

-adam

By Adam Goucher on May 3, 2008

@greg: Overall, Zed is pretty much on the money. Managers tend to make decisions which are outside the scope of their competence, and then hold the technical staff responsible for the failures. This is why I’m my own boss! There is indeed a grain of truth in the Dilbert strips.

For an insight from the larger sphere of high-tech engineering take a look at the disaster of the space shuttle Challenger. If you don’t have time for the whole Rogers Commission report, here’s an extract from the Wikipedia article:

“”"NASA managers had known that contractor Morton Thiokol’s design of the SRBs contained a potentially catastrophic flaw in the O-rings since 1977, but they failed to address it properly. They also ignored warnings from engineers about the dangers of launching on such a cold day and had failed to adequately report these technical concerns to their superiors.”"”

Managers are driven by other issues than technical staffers, and tend to make their decisions based on other criteria than the technical. The engineers had addressed the problem as well as they were able by reporting the issue, but it wasn’t up to them whether the launch proceeded.

By Steve Holden on May 3, 2008

“Compare you answer to what you’d say if an electrical engineer designed a cell phone that would occasionally connect a 911 call incorrectly.”

The funny thing is that my cellphone DOES do this. When it’s on keypad lock, it allows the keys 9 - 1 - 1 - TALK be pressed in succession. My guess is that it allows it to be dialed when someone doesn’t know how to unlock the keypad. Unfortunately, I’ve accidentally dialed 911 twice since I got the phone…

By Dave Cooper on May 3, 2008

To what degree are the developers responsible/accountable? It likely depends on the laws surrounding product liability [1]. I’d be surprised if the individual software developers were actually prosecuted. Take the Ford Pinto as an example [2]. The engineering team produced a car with a design defect, but I don’t think any of them were held criminally liable—Ford Motors Inc. was.

And maybe the Internet phone provider knew that their software would not handle 911 calls correctly. Skype, for example, “does not intend to support emergency calls” [3]. I have a couple of problems with this. First: I don’t think I’ve ever read—nor ever will read—and EULA. Second: I wouldn’t be very happy if my cell phone’s 911 TOS had the clause “your results may vary.”

At the end of the day, I would have to ask myself: would I be able to sleep at night, knowing that software I had written caused even a single death (regardless of the legal implications)? Likely not; and remember: I was born without the shame gene.

[1] http://en.wikipedia.org/wiki/Product_liability#Products_Liability_and_Strict_Liability
[2] http://en.wikipedia.org/wiki/Ford_Pinto#Safety_problems
[3] http://www.skype.com/legal/eula/

By Miles Thibault on May 5, 2008