| || |
I currently am reviewing the symptoms we use (and used) to classify our defects.
Over the years, we have slowly soiled our own nest and a cleanup of symptoms is in order.
Here's the list of used symptoms:
Crash - Data Loss
Crash - No Data Loss
Display Column Width Small
Incorrect Data Presentation
Incorrect Error Message
JCL Error/Run Instructions
Training Attention Req'd
Web Design Issue
Web Page Design Issue
WEB Standard Compliance
Many of them are not even symptoms and then there are duplicates and other problems with it as well. So we are ready for a clean up.
By the way, we currently have a repository of over 25,000 defects logged over the last 5 years and we just migrated this all to a new defect tracking tool. (TestTrack Pro)
I tried to find solid symptom lists on the internet and the discussion forums but no luck there.
We basically want to divvy the symptoms up in following categories:
For instance we could have the following symptoms for Database:
Database: Missing Table
Now my question is: Do you have a comprehensive list of defect symptoms or some advice on a solid classification hierarchy?
Re: Defect symptoms
Just some thoughts:
According to my experience, you have a list of "symptoms", which are not all symptoms. Some of them suppose to be in completely other fields.
I suppose that will be easier to clean this list if we know what other fields do you use to clasify a bug:
- do you have them clasified by Functional/performance/security/... ?
- do you have them clasified by severity ?
- do you have them by the development cycle it was noticed ?
The categories you've proposed , don't seem very logical to me:
- you could have database requirements bugs... and you'll have. Don't let the tester decide in which category to clasify it.
-the list (of very probable combinations) could continue.
Re: Defect symptoms
***According to my experience, you have a list of "symptoms", which are not all symptoms. Some of them suppose to be in completely other fields.***
Yes exactly! That's what I said. That's the reason this list needs an overhaul.
We have symptom, severity, priority, product name, release found, component, testevent found, config info etc. So we are already collecting a lot of detail.
The whole idea of creating top level categories is to make the classification more structured.
For instance we would get
* Database: Missing Table
* Application: Unexpected behavior
* Data: Corruption
* Standards: GUI standard infraction
We actually can accommodate a lot of symptoms, so I feel we are not limiting ourselves in any way or form.
Currently we use a subset of our symptoms to set a default severity. For instance "Crash" results in severity "High", "Spelling error" in low etc.. With all the noise now in our symptoms list, the automatic severity setting does not work consistently.
Anyway, thanks for your feedback.