|
|
 | | From: | William | | Subject: | Re: Every software defect should be able to reproduce?? | | Date: | Mon, 10 Jan 2005 10:14:53 -0600 |
|
|
 | wrote in message
news:1105318437.384248.112490@c13g2000cwb.googlegroups.com...
> If the testers find a bug (unexpected behavior) in software, but that
> bug cannot be reproduced every single time for the same input,
Well, if you could guarantee the same input, you could always reproduce
it - unfortunately, what most people think of as input is only a small part
of the real input to the program.
> or maybe
> because of the environment setting problem, we may consider it as an
> intermittent bug? But we should still consider this as a bug?
Yes, but it's always possible you'll never be able to fix it. I've seen bugs
live through several releases of a product because no one could reprodce
them despite considerable effort. Eventually, though, someone usually
does manage to capture enough information about the problem so it can
either be reproduced, or, at least, the events leading to it can be
identified
and corrected.
I think the real test is not if you can reproduce it, but if users ever
reproduce it.
The only time I wouldn't consider something like this *my* bug is if it
is the result of a third-party's mistake. (I had an install project fail for
no obvious reason once - turned out to be a bad patch from MS that
affected installs written in an odd, but perfectly legal, way.) -Wm
|
|
|
