Comments on: Where are the dead bodies?

By: TemporalBeing

TemporalBeing — Mon, 30 Nov 2009 17:59:45 +0000

Well…I’ll guess two things:

1) There are likely hardware (e.g. non-software) limits in place that sufficiently overcome the faults of software in most cases.
2) Deaths probably get related to hardware issues (no limiter in place to keep the software from going there) than getting reported as software.

Also, don’t forget that while something may keep someone from being killed there is also the issue of injury – regardless of degree. How much does software related faults contribute to injuries? Probably even harder to tell.

By: Whither Software Engineering « Blogs are like opinions. Everybody has one…

Thu, 26 Nov 2009 11:56:43 +0000

[...] extensive professional analysis of any large software failure”. Meyer is not alone. “Where are the dead bodies?” asks Derek M. Jones who also writes: “The lack of dead bodies attributed to a software [...]

By: Pascal Cuoq

Pascal Cuoq — Wed, 18 Nov 2009 23:34:31 +0000

The Patriot 0.1s bug is also often cited, and certainly at least one life was lost that could have been saved — but this was war. Or maybe we hear a lot about it because we have colleagues who have a numerical precision analysis tool they want to sell you, although in this case, a simple compiler warning that a floating-point literal cannot be represented exactly, properly understood, would have been enough to avoid embarrassment.

Generally speaking, it is true that for instance, research in the field of software safety uses the same examples again and again to justify its existence, even when use of the methods they advocate remains — as of 2009 — marginal.

On the other hand, software is deterministic and often replicated exactly into thousands of devices. Perhaps it is feared — and perhaps even with reason — the death toll would become immediately unacceptably high if the slightest compromise in safety was allowed.

By: Twitted by frama_c

Twitted by frama_c — Wed, 18 Nov 2009 23:15:08 +0000

[...] This post was Twitted by frama_c [...]

By: Steven Oliver

Steven Oliver — Wed, 18 Nov 2009 17:38:40 +0000

I don’t know if you looked or not, and I don’t know if they’d be included in the US National Transportation Safety Board’s database, but have you looked into military deaths related to software? I know at least pilot died the new F-22 stealth fighter for Lockheed Martin. How much of that death was software related I don’t know though.