Closing bugs (?)

Closing bugs (?)

[Sven Wischnowsky]

Sourceforge supports the `fixed' and `closed' states. Hm, do we want
to leave it to one of the administrators to actually close a bug or
should the person who fixed (or tried to fix) it do that?

And why this distinction?

Bye
 Sven

P.S.: 103797, 103798

--
Sven Wischnowsky                         wischnow@informatik.hu-berlin.de

Re: Closing bugs (?)

[Bart Schaefer]

On Apr 4,  4:33pm, Sven Wischnowsky wrote:
} Subject: Closing bugs (?)
}
} Sourceforge supports the `fixed' and `closed' states. Hm, do we want
} to leave it to one of the administrators to actually close a bug or
} should the person who fixed (or tried to fix) it do that?
} 
} And why this distinction?

A bug can be closed without being fixed, i.e. "that's not a bug, it's a
feature," or "seeming bug was caused by pilot error," etc.

The way I've typically handled it with GNATS before is that the person
who fixes the bug changes the state to "fixed" ("feedback" in GNATS),
and then it's up to the administrator and/or the person who reported
the bug to agree that it's fixed and change it to "closed".

But maybe we don't need that much supervision, and maybe it's OK to
leave a bug in the "fixed" state forever.

-- 
Bart Schaefer                                 Brass Lantern Enterprises
http://www.well.com/user/barts              http://www.brasslantern.com

RE: Closing bugs (?)

[John Grossi]

> On Apr 4,  4:33pm, Sven Wischnowsky wrote:
> } Subject: Closing bugs (?)
> }
> } Sourceforge supports the `fixed' and `closed' states. Hm, do we want
> } to leave it to one of the administrators to actually close a bug or
> } should the person who fixed (or tried to fix) it do that?
> } 
> } And why this distinction?

Speaking as a QA person, there is a large distinction between closed and
fixed states.

A "fixed bug" is one that the developer has entered code that he believes
fixes the problem.

A closed bug is one that's been built, and successfully survives a unit test
and system test.

Lots of code crosses my desk that is "fixed" and somehow doesn't work because
of that piece of kryptonite in my pocket.

-John

Software QA Engineer
GTE Internetworking
Burlington, MA, USA

Re: Closing bugs (?)

[Sven Wischnowsky]

Bart Schaefer wrote:

> On Apr 4,  4:33pm, Sven Wischnowsky wrote:
> } Subject: Closing bugs (?)
> }
> } Sourceforge supports the `fixed' and `closed' states. Hm, do we want
> } to leave it to one of the administrators to actually close a bug or
> } should the person who fixed (or tried to fix) it do that?
> } 
> } And why this distinction?
> 
> A bug can be closed without being fixed, i.e. "that's not a bug, it's a
> feature," or "seeming bug was caused by pilot error," etc.
> 
> The way I've typically handled it with GNATS before is that the person
> who fixes the bug changes the state to "fixed" ("feedback" in GNATS),
> and then it's up to the administrator and/or the person who reported
> the bug to agree that it's fixed and change it to "closed".
> 
> But maybe we don't need that much supervision, and maybe it's OK to
> leave a bug in the "fixed" state forever.

Hm. In this ugly Web-Interface we have (in the list), the state isn't
shown, so you can only tell if a bug is (supposed to be) fixed after
clicking on it etc. I think it would be nice to keep the list of open
bugs small -- but of course that's only really a problem when there
are more than there are now.

Bye
 Sven


--
Sven Wischnowsky                         wischnow@informatik.hu-berlin.de