ISV software quality; tortology or oxymoron…

on Dec 10, 08 • by Gwyn Fisher • with No Comments

It’s kind of bizarre, but in my pre-Klocwork experience of running ISV development groups, from small teams to global enterprises, it never struck me as wrong that we would routinely ship software containing critical bugs. We knew we were doing it. We knew, on some abstract underground never-to-be-admitted layer...

Home » Software Quality » ISV software quality; tortology or oxymoron…

It’s kind of bizarre, but in my pre-Klocwork experience of running ISV development groups, from small teams to global enterprises, it never struck me as wrong that we would routinely ship software containing critical bugs. We knew we were doing it. We knew, on some abstract underground never-to-be-admitted layer of our deepest darkest souls, that this was a “bad thing.”

But mostly, we knew that when somebody found a bug we could just send them a patch.

And what’s more, we knew that customers expected this behavior. We got requests to “send us a patch quickly, this is causing me {insert unpleasant business scenario here}.” We had customers telling us that we were a great company to deal with, because we responded to patch requests quickly.

We weren’t the bad kids on the block, we were right up there with the upstanding corporate citizens of the software world!

Why are software companies allowed, perhaps even expected, to mess up on this scale? If we were engineers, we’d be sued. If we were doctors, we’d be in court before you could say something snappy and relevant like “most software sucks.” But for ISV developers, hey, it’s just business.

Talk to developers in the embedded space, or mil/aero, or telecoms, or a bunch of other places where this approach to business isn’t acceptable and it’s like talking to engineers. They understand what their tolerances are, they understand what it takes to make quality software, and they fundamentally understand on a basic level what happens to them and their company if they get it wrong. Inventory turnover is one thing, killing people is something else.

So is it just that ISV software is unimportant? That ISV software won’t ever be put in a situation where it could be life-threatening (or life-enhancing), so it doesn’t matter if it crashes all the time?

Obviously that’s not the case. Software created by ISVs is used all over the world, in all manner of situations, some dire, some not, but at the end of the day, in every situation, there’s a name on that software.

Your name.

Related Posts

Leave a Reply

Your email address will not be published. Required fields are marked *

Scroll to top