Policies on old versions?

Marcus Denker denker at iam.unibe.ch
Mon Dec 12 13:11:12 UTC 2005


Hello,


On 12.12.2005, at 07:26, Cees De Groot wrote:

> I'm looking at Mantis #1460 which is a bug filed against 3.7. What
> should we do with these old bugs that don't have a simple test to
> reproduce?
> - Reject them?
> - Ask the submitter to try to reproduce on 3.8?
> - Load a 3.7 image, reproduce, if yes load a 3.8 image, try to  
> reproduce, ...?
> Clearly, I think the third option would put quite a burden on the bug
> handlers (and thus slow down harvesting). As the user invested some
> time to compose a bug report, is it too much to ask him to retry this
> on a later version?

Asking the submitter is fine. We have a lot of bug reports / fixes,  
and we
should make sure that we don't put too much time into the bad reports.
So for a report that is somewhat defective, I vote for asking the  
submitter
to fix it. If the report is old, I'd even close is with an  
explanation to resubmit.

Overall, I'd like to have a bug and fix tracking system that is more  
workflow-based
(or has a workflow view). For the harvesters, it would be nice to not  
see such
items that are delegated back to the submitter...

       Marcus



More information about the Squeak-dev mailing list