[Mapserver-dev] Bug status needs attention

[Daniel Morissette]

Sean Gillies wrote:
> 
> My immediate concern is generating an authoritative list of known bugs  
> in 4.2.4
> without having to read and verify each bug, and email reporter and  
> owner for
> clarification.
> 
> Below is a query on unresolved (minus enhancement requests) bugs which  
> have a
> version marked 4.2:
> 
> http://mapserver.gis.umn.edu/bugs/buglist.cgi? 
> short_desc_type=allwordssubstr&short_desc=&product=MapServer&version=4.2 
> &long_desc_type=allwordssubstr&long_desc=&bug_file_loc_type=allwordssubs 
> tr&bug_file_loc=&keywords_type=allwords&keywords=&bug_status=NEW&bug_sta 
> tus=ASSIGNED&bug_status=REOPENED&bug_severity=blocker&bug_severity=criti 
> cal&bug_severity=major&bug_severity=normal&bug_severity=minor&bug_severi 
> ty=trivial&emailassigned_to1=1&emailtype1=substring&email1=&emailassigne 
> d_to2=1&emailreporter2=1&emailcc2=1&emailtype2=substring&email2=&bugidty 
> pe=include&bug_id=&votes=&changedin=&chfieldfrom=&chfieldto=Now&chfieldv 
> alue=&cmdtype=doit&namedcmd=Active+1+Day&newqueryname=&order=Reuse+same+ 
> sort+as+last+time&field0-0-0=noop&type0-0-0=noop&value0-0-0=
> 
> Could you help me rally developers to go through this list and, where  
> appropriate,
> change status of bugs to FIXED, WONTFIX, or UNCONFIRMED?  Afterwards,  
> I'll report
> each remaining unresolved bug as a known bug whether status is NEW or  
> ACCEPTED.
> 
> Once 4.2.4 is released, I'll think a bit more about how to better do  
> this for
> the 4.4.0 release, and about the larger issues.
> 

Sean,

I'm trying to do what you suggest with the list of 4.2 bugs assigned to 
me or other DMSG staff and that doesn't work. There are several bugs 
that will require some work/research before we can decide if they are 
really bugs or what we do with them. I guess they are those that should 
be marked as unconfirmed in your plan, but how do you that? I have never 
seen an option to mark bugs confirmed or not.

I think it is a legitimate objective to produce a list of known bugs for 
4.2, but I'm not convinced that a complete list of open bugs as you 
suggest has much real value for the users. I don't think we should list 
every minor bug in a known bug list or we just risk scaring the users 
with a large number of bugs when the software is actually very stable 
for 99.5% of users. Most open bugs that haven't been addressed yet are 
enhancements or stuff that users have very few chances of running into. 
Only important bugs that remain unfixed in 4.2 should be mentioned. At 
the moment I am not aware of any, are you?

Also, I think that your current bugzilla query has chances of missing 
important 4.2 bugs that were fixed only in 4.3 and that could possibly 
have been of interest to the users (more than the 30 other minor bugs 
that the query returns). Those bug will remain in 4.2 but won't show up 
in your query. This happens often for minor bugs with a non-trivial fix 
that we don't backport because we don't want to risk breaking the stable 
4.2 branch.

Anyway, I'd like to help with your buglist but I'm not sure how. I would 
propose that we go ahead with the 4.2.4 release and get back to the list 
of open bugs when we have a better plan.

Daniel
-- 
------------------------------------------------------------
  Daniel Morissette               dmorissette at dmsolutions.ca
  DM Solutions Group              http://www.dmsolutions.ca/
------------------------------------------------------------