[gmx-developers] bug in 3.2.x not fixed in 3.3.1

[Erik Lindahl]

Hi,

On May 9, 2009, at 9:24 AM, hessb at mpip-mainz.mpg.de wrote:
>
> I don't know if commenting is the issue here.
>
> A bug should simply never be resolved to status "fixed"
> when a patch has not been committed to at least the head branch.

Agreed - the last couple of years I think we've been better at closing  
bugs with a comment "committed both to release and head branch".

Fortunately I think this is the first time a bug has been formally  
closed but the patch still not applied - our apologies!


The best solution is probably to make better use of Bugzilla's bug  
status field: http://bugzilla.gromacs.org/page.cgi?id=fields.html#status

Right now we only use new/unconfirmed/resolved in practice. I would  
propose:

1) *ALL* bugs are UNCONFIRMED when users report them.
2) Once they have been confirmed by somebody else, it's "NEW"

3) Once a proposed patch/solution is available, we move it to  
"RESOLVED".
4) Then we should only move tickets to status "CLOSED" when we have  
confirmed the fix is in all branches of CVS...


The first thing we'll need to do is to go through past bugs and check  
if they indeed should be closed, but most of them will be trivial I  
think.

(And then users should check that their bug gets "CLOSED", not only  
"RESOLVED"!)


So, considering this bug is technically still only "RESOLVED", can we  
move it to "CLOSED" ? ;-)

Cheers,

Erik