Re: Bug in plotting - 3.0.4 RC1

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: Bug in plotting - 3.0.4 RC1

John W. Eaton
Administrator
[this part of the discussion moved from the bugs list.  --jwe]

On 23-Jan-2009, John W. Eaton wrote:

| On 23-Jan-2009, John W. Eaton wrote:
|
| | If you are talking about the change from root_figure to root, then
| | it looks to me like that change has already been made.
|
| Oops, I was looking at the wrong sources.  But is it necessary to make
| this change for 3.0.x?  Does it cause any other trouble if it remains
| "root figure"?  That is not a regression from previous 3.0.x releases.
|
| I do see that the legend appears to be reversed.  I guess that is a
| consequence of switching the order of handles.

May I suggest that in the future we restrict ourselves to fixing only
regressions in the stable releases?  What I mean is that if the bug is
new, or was present in a previous stable release, we would fix it in
the main branch but NOT in the stable branch.  Although this may be
upsetting to some who report bugs in (say) 3.0.3 and then find that
the problems are not fixed in 3.0.4, I think that attempting to fix
bugs that are not regressions tends to move us away from the stability
we are promising for the stable release series.

Comments?

jwe
Reply | Threaded
Open this post in threaded view
|

Re: Bug in plotting - 3.0.4 RC1

Søren Hauberg
fre, 23 01 2009 kl. 10:52 -0500, skrev John W. Eaton:

> [this part of the discussion moved from the bugs list.  --jwe]
>
> On 23-Jan-2009, John W. Eaton wrote:
>
> | On 23-Jan-2009, John W. Eaton wrote:
> |
> | | If you are talking about the change from root_figure to root, then
> | | it looks to me like that change has already been made.
> |
> | Oops, I was looking at the wrong sources.  But is it necessary to make
> | this change for 3.0.x?  Does it cause any other trouble if it remains
> | "root figure"?  That is not a regression from previous 3.0.x releases.
> |
> | I do see that the legend appears to be reversed.  I guess that is a
> | consequence of switching the order of handles.
>
> May I suggest that in the future we restrict ourselves to fixing only
> regressions in the stable releases?  What I mean is that if the bug is
> new, or was present in a previous stable release, we would fix it in
> the main branch but NOT in the stable branch.  Although this may be
> upsetting to some who report bugs in (say) 3.0.3 and then find that
> the problems are not fixed in 3.0.4, I think that attempting to fix
> bugs that are not regressions tends to move us away from the stability
> we are promising for the stable release series.
>
> Comments?

If this is not done, then I think it's hard to tell the difference
between the stable and the main branch. So, I agree.

Søren