Thanks for responding, Dag. A couple comments inline...
Dag H. Wanvik wrote:
Rick Hillegas <[EMAIL PROTECTED]> writes:
I'm going through our outstanding bugs now and finding that I would
like to divide them into the following buckets:
I think something that can make us more efficient in categorizing bugs
and retaining the result of such analyses longer is useful, cf. the
list we keep for high value fixes. Maintaining manual lists is
tedious, though.
We already use the category fields for tagging "newcomer" issues, so
why not augment that with suitable tags? Since an issue can have many
categories it seems suited to tagging. I guess the flags field is a
good candidate location too - "regression" seem to be a similar tag -
but maybe flags are harder to add?
Just to be precise, I think that you're referring to the drop-down
"Component" menu. I agree that whatever the original meaning of this
menu, it has been extended to include items like "Newcomer", "Security",
and "Performance", which don't fit my definition of a product Component.
If it's ok to continue overloading this menu, then I agree that this is
an attractive place to list these bug categories.
Data Corruptions
Crashes
Wrong Results
+1
Security Holes
We already have a security category. But I guess we could use this one
as well? +0.5
Agreed.
Slowness
+1
This one I think we already have today too under the name "Performance".
Exceptions
Not sure what this one means..
Right. This is a bad name. What I mean here are failures other than
crashes or corruptions. E.g., compiler exceptions, out-of-memory errors,
aborts during execution. What would be a better name for this: Aborts?
Other bugs
Do we need that category?
I thought it would be useful to partition bugs among these categories
for easy filtering via JIRA. I don't think you can write a JIRA filter
whose sense is "none of the above".
Other candidates:
- standard deviation
- drivers differ
Now I have to ask you for a little more context! What do these
categories mean?
Thanks,
-Rick
