| Peter Belt 2006-08-28, 8:05 am |
| > Qingpei Huwrote:
>
>
> - Detection does not stop if correction didnot.
>
> Why not?
> Any practical limitation?
>
>
Yes, the last (?) correction could be either fatal or the solution to
all problems. You really want to know. So after the last correction
you need 1 more round of testing.
> - if you classify your findings, (Showstopper- High- Medium- Low-
> Presentation) you can decide to stop correction step by step.
> First you stop fixing the issues with severity Low, then you stop
the
> Medium's, then you stop the High's. You can continue fixing the
> Presentation issues (Cosmetics) if they do not have any impact on
> functionality. Showstoppers can be evaluated like this: if it was
> detected 1 w after the release date, would it result in a
hotfix?
> if yes, you'd better fix it immediately. If not, it's probably not
> worth fixing and testing it right now.
>
> But how do you know when to stop?
>
>
> Thanks,
>
> Neon
I can't tell you, but your organization can, because it depends on:
- required time to market
- required level of quality
- quality state of the project
There is not one moment where you stop fixing, there are several
moments where you stop fixing certain types of issues.
Generally, it's wise to stop fixing issues as soon as possible when
the risks they introduce are bigger then the profit of the fix.
That's why you need to classify the reported problems on 'severity',
and need to classify the solution on 'impact'.
Regards,
Peter
Posted from the Dutch software testing community at www.testforum.nl
|