[Zope3-dev] The bug fixing problem

[Martijn Faassen]

Jim Fulton wrote:
> 
> On Jul 11, 2006, at 8:05 AM, Patrick Gerken wrote:
> ...
>> As a non committer I would like to note that it was easy for me to
>> search if somebody already submitted a bug I found, and submit a new
>> patch, it was also trivial to add the for the bugfix and the test. The
>> only thing which IS still not easy is to find a way to see that patch
>> appear in Zope itself.
>> I marked the bug as bug + bugfix but nobody cares. That is much more
>> discouraging than what I can not do nice wiki links to in my bugreport
>> other bugtracker items or svn sources like it is possible in trac
>> itself.
> 
> Well said.  There are people who care.  A few of us have biin working 
> through
> the bugs.  I wish more people were. (I'm stuck on a fairly hairy one 
> myself.)
> 
> Regardless of what tool we use, we need to be committed to not
> letting bug reports languish in the collector.
> 
> Personally, I'd like to have a feature freeze until we've cleane dup
> *all* of the outstanding bugs, not just the critical ones.

While I'm not against a period of time where we're committed to cleaning 
out bugs from the tracker, I find myself wondering whether that won't 
mean an indefinite feature freeze... Will that really get people more 
motivated to fix bugs?

Better information for both developers and bug submitters and a better 
way to defer bugs for following releases might give us more flexibility 
and a better ability to select the issues we consider truly important 
and might increase our collective memory.

The other suggestion I made elsewhere is the ability for developers to 
add breaking tests to the codebase (explicitly marked as such, and not 
normally run). This might make it easier for people who got halfway 
fixing a bug to let their knowledge not be lost, and might also make it 
easier for people to find bugs to fix (that already have tests!).

Regards,

Martijn