[GRASS-dev] Re: GRASS bugs

[Otto Dassau]

Am Mittwoch, 2. August 2006 19:15 schrieb Maciej Sieczka:
> Glynn Clements napisa?(a):
> > Moritz Lennert wrote:
> >>> But generally: why wait for a bugfix party? Just go ahead...
> >>
> >> I don't think it is about waiting, it is about creating a special,
> >> motivating event where everyone works together on one goal.
> >
> > The one thing which would do the most to motivate me to fix bugs would
> > be a list of clearly-identified, verified, reproducible bugs.
>
> I can do that and will send it to the list. Most likely next week. If
> there is no obvious test case I will try to create it based on spearfish.
>
> To all you Guys who were willing to help with testing during the bug
> sqaushing party: please give me a hand with that. Each of you, pick few
> reports, verify it using spearfish, or, for bugs in the interface etc.,
> just verify if it is still present in current CVS. Then 'reply' from the
> tracker CCing me at my email. You can do it all even as guest, just make
> sure you sign with name and email. If I counted OK there should be 5
> testers alltogether. Such a band can clean the mess up easily. Anybody -
> please join!

Hi,

would it make sense to create a list of bug IDs to verify - maybe in wiki? 
People can add themselves together with their favorite IDs to make sure that 
bugs won't be verified twice. 

regards,
  Otto

> > Back when I used to check the RT occasionally, I would generally skip
> > any report where I couldn't quickly reproduce the problem from the
> > information contained in the report.
>
> I agree this is often the case that people don't document their reports
> well enough, but this doesn't mean such reports should be ignored
> completely.
>
> > As this tends to be the case for the vast majority of reports, I
> > eventually gave up checking the RT altogether
> > (other than to kill requests generated by spam email to the grass-bugs
> > address).
>
> Please don't waste your time. This is such an easy task I can do it
> myself (and I'm doing it).
>
> > If people want a bug to get fixed, by far the most important factor is
> > providing a clear mechanism to demonstrate the bug, either using the
> > spearfish dataset or starting from a new location (i.e. using
> > self-generated test data).
>
> I know, I've failed to do it often too. But I also reported bugs, and
> saw other's reports, documented well but still never taken care of.
>
> You are right there is a need to clean up the tracker further to make
> developers' work easier. I'll work on this as I can.
>
> Maciek