RESOLVED states in bugzilla

RESOLVED states in bugzilla

[Wolfgang Bangerth]

Bugzilla presently offers lots of resolutions for RESOLVED PRs. These I 
think are useful:
    Fixed
    Invalid
    Duplicate
These are doubtful. I haven't seen a PR where someone said "We're never 
going to fix this", but it won't hurt to have this; and WORKSFORME is not 
a very strong statement:
    Wontfix
    Worksforme

These I think are not useful:
    Remind
    Later
If a bug is not fixed now, it should remain open, not be closed and have a 
special state. Dan, can you say who invented these states and why? I don't 
see any use for them, and by the rule of maximal simplicity, would like to 
get rid of them if possible.

W.

-------------------------------------------------------------------------
Wolfgang Bangerth              email:            bangerth@ices.utexas.edu
                               www: http://www.ices.utexas.edu/~bangerth/

Re: RESOLVED states in bugzilla

[Daniel Berlin]

On Sat, 31 May 2003, Wolfgang Bangerth wrote:

>
> Bugzilla presently offers lots of resolutions for RESOLVED PRs. These I
> think are useful:
>     Fixed
>     Invalid
>     Duplicate
> These are doubtful. I haven't seen a PR where someone said "We're never
> going to fix this", but it won't hurt to have this; and WORKSFORME is not
> a very strong statement:
>     Wontfix
>     Worksforme
>
> These I think are not useful:
>     Remind
>     Later
> If a bug is not fixed now, it should remain open, not be closed and have a
> special state. Dan, can you say who invented these states and why?

Built into Bugzilla.
> I don't
> see any use for them, and by the rule of maximal simplicity, would like to
> get rid of them if possible.

I'll hide those two immediately.

Did we ever come to consesus on whether we wanted VERIFIED and CLOSED?
>
> W.
>
> -------------------------------------------------------------------------
> Wolfgang Bangerth              email:            bangerth@ices.utexas.edu
>                                www: http://www.ices.utexas.edu/~bangerth/
>
>
>

Re: RESOLVED states in bugzilla

[Wolfgang Bangerth]

> Did we ever come to consesus on whether we wanted VERIFIED and CLOSED?

At least I didn't see any strong _pratical_ arguments. There might be some 
theoretical reasons, but I don't think they would stand reality.

I think if nobody comes forward with a good plan
- how to make use of them
- AND HOW IS GOING TO DO THE WORK (please no obscure cron jobs)
in the next few days, go ahead and get rid of VERIFIED. I'm willing to 
take the heat of this is the wrong decision.

W.

-------------------------------------------------------------------------
Wolfgang Bangerth              email:            bangerth@ices.utexas.edu
                               www: http://www.ices.utexas.edu/~bangerth/

Re: RESOLVED states in bugzilla

[Gerald Pfeifer]

On Sat, 31 May 2003, Wolfgang Bangerth wrote:
>> Did we ever come to consesus on whether we wanted VERIFIED and CLOSED?
> I think if nobody comes forward with a good plan
> - how to make use of them
> - AND HOW IS GOING TO DO THE WORK (please no obscure cron jobs)
> in the next few days, go ahead and get rid of VERIFIED. I'm willing to
> take the heat of this is the wrong decision.

Be my guest to send part of the flames in my direction. ;-)  I fully agree
with your proposal.

KISS.

Gerald
-- 
Gerald "Jerry"   pfeifer@dbai.tuwien.ac.at   http://www.pfeifer.com/gerald/

Re: RESOLVED states in bugzilla

[Daniel Berlin]

On Saturday, May 31, 2003, at 11:08  PM, Wolfgang Bangerth wrote:

>
>> Did we ever come to consesus on whether we wanted VERIFIED and CLOSED?
>
> At least I didn't see any strong _pratical_ arguments. There might be  
> some
> theoretical reasons, but I don't think they would stand reality.
>
> I think if nobody comes forward with a good plan
> - how to make use of them
> - AND HOW IS GOING TO DO THE WORK (please no obscure cron jobs)
> in the next few days, go ahead and get rid of VERIFIED. I'm willing to
> take the heat of this is the wrong decision.
>

I'll deal with verified and closed in a moment, but LATER and REMIND  
are now gone.

> W.
>
> ----------------------------------------------------------------------- 
> --
> Wolfgang Bangerth              email:             
> bangerth@ices.utexas.edu
>                                www:  
> http://www.ices.utexas.edu/~bangerth/
>
>

Re: RESOLVED states in bugzilla

[Joseph S. Myers]

On Sun, 1 Jun 2003, Daniel Berlin wrote:

> I'll deal with verified and closed in a moment, but LATER and REMIND  
> are now gone.

Bug 11034 (the one that was marked as LATER) still shows as RESOLVED, with
the Resolution field now blank.

-- 
Joseph S. Myers
jsm28@cam.ac.uk

Re: RESOLVED states in bugzilla

[Daniel Berlin]

On Sunday, June 1, 2003, at 12:49  PM, Joseph S. Myers wrote:

> On Sun, 1 Jun 2003, Daniel Berlin wrote:
>
>> I'll deal with verified and closed in a moment, but LATER and REMIND
>> are now gone.
>
> Bug 11034 (the one that was marked as LATER) still shows as RESOLVED, 
> with
> the Resolution field now blank.
I just set it to SUSPENDED.

Re: RESOLVED states in bugzilla

[Wolfgang Bangerth]

> I'll deal with verified and closed in a moment, but LATER and REMIND  
> are now gone.

Thanks a lot!

W.

-------------------------------------------------------------------------
Wolfgang Bangerth              email:            bangerth@ices.utexas.edu
                               www: http://www.ices.utexas.edu/~bangerth/

Re: RESOLVED states in bugzilla

[Joseph S. Myers]

On Sat, 31 May 2003, Wolfgang Bangerth wrote:

> These are doubtful. I haven't seen a PR where someone said "We're never 
> going to fix this", but it won't hurt to have this; and WORKSFORME is not 
> a very strong statement:
>     Wontfix
>     Worksforme

I dislike WONTFIX; I think for all the bugs closed that way, INVALID is
more appropriate (i.e., declare it to be not a GCC bug rather than to be a
GCC bug that won't be fixed).  WORKSFORME is more reasonable.

> These I think are not useful:
>     Remind
>     Later
> If a bug is not fixed now, it should remain open, not be closed and have a 
> special state. Dan, can you say who invented these states and why? I don't 

Yes, we created SUSPENDED and made sure that it counts as open in default 
searches for this purpose.  The one bug closed as LATER should be open as 
SUSPENDED instead.

-- 
Joseph S. Myers
jsm28@cam.ac.uk

Re: RESOLVED states in bugzilla

[Gareth McCaughan]

Wolfgang Bangerth wrote:

> These are doubtful. I haven't seen a PR where someone said "We're never
> going to fix this", but it won't hurt to have this; and WORKSFORME is not
> a very strong statement:

WORKSFORME has a meaning that seems to me strong enough to be
useful. (And not equivalent to INVALID as someone else suggested.)
It means:

    At least one of us has looked at this; none of us
    has been able to reproduce the buggy behaviour.
    If we could, we'd agree that it's a bug.

Whereas INVALID means:

    The behaviour you report is not a bug.

I think that's a useful distinction. For instance: suppose
a new bug report, just like the already resolved one, comes in.
For an INVALID bug you just file it as a duplicate and forget
about it. For a WORKSFORME bug you should consider seriously
that maybe it's more reproducible than you thought.

-- 
g