On Nov 7, 2015, at 3:39 PM, Moose Finklestein <[email protected]> wrote:

> The article goes on to pull out data breaches and other failures,
> insisting that things like this would never happen in physical
> engineering. Wishful thinking.  How many bridges failed due to poor
> design?  How many buildings have collapsed?

I agree that the arguments weren’t great, but if I may play devil’s advocate 
for a moment…

What happened in the case when those buildings collapsed or those bridges 
failed?  Or, to use your later examples, when laptops had a memory board 
failure or caught on fire?

In those cases, someone was likely held responsible.  If something broke due to 
an unforeseen materials problem or a never-before-encountered wind shear, then 
there was a probably a nice journal article written about it.  However, if the 
engineer in charge cut corners and should have known better, there was likely 
some liability attached.  If your laptops are defective, you get new ones (or 
free repairs).  If you don’t, there’s usually a class-action lawsuit.  
Professional engineers need to sign off on the components they use or recommend 
for a design.  If they recommend the wrong ones and something goes south, 
they’re in big trouble.

Compare this to software flaws in most applications (I’m excluding the space 
shuttle and other exceptional programs here).  If something breaks, there’s 
usually “no warranty”.  What if all my work gets destroyed by a software bug 
(it’s certainly happened to me, with some expensive software)?  Is there a 
class-action lawsuit? There are critical software bugs fixed every day all year 
long, but it’s just accepted that it’s part of the deal.  If you (as a 
customer) called up someone like Microsoft and yelled at them about some work 
you’d lost due to Word crashing, you’d get some sympathy and not much else.

How do you know that the software you’re buying has had some reasonable 
standard of care put into its development?  Without access to the source, you 
only have a particular company’s reputation to go on.

Note that I don’t necessarily believe we should license everyone who writes 
software for a living.  I particularly feel that open-source software should be 
able to disclaim some liability (no warranty for any purpose, etc, etc) since 
liability could really hurt people wanting to release stuff.

But if there were at least an option to certify that software follows some best 
practices, that would be helpful.  Of course, that requires some kind of 
magical, all-encompassing, difficult-to-fake, non-burdensome yet worthwhile 
certification process that doesn’t just become a bureaucratic box to check on a 
project.

I suppose companies could just start offering some kind of performance warranty 
on their work voluntarily and hope people are willing to pay a little extra for 
the assurance.  However, since most people seem to accept software bugs as par 
for the course, I’m not sure who would go out of their way to do that.

Jason

_______________________________________________
Discuss mailing list
[email protected]
https://lists.lopsa.org/cgi-bin/mailman/listinfo/discuss
This list provided by the League of Professional System Administrators
 http://lopsa.org/

Reply via email to