On Friday 15 May 2009 21:38:04 Sven Neumann wrote:

> Keeping the pixel under the mouse cursor fixed has the advantage
> that the behavior for zooming in and out is consistent. 

Well, it would be consistent if gimp consistently kept this 
pixel-cursor relation. But it is not (gimp 2.6.6). Open any image 
smaller than window, point out on anything not in center. Zoom in. 
Voila, you are pointing at something else now, because gimp does not 
shift view of the image.

> If you zoom 
> out too far, you can easily zoom back in without loosing the area
> of interest.

Only if image does not fit the window (gimp 2.6.6). If SVN still does 
this it is a bug then, but I assume it is or will be fixed.

> I think that clearly outweights the lack of context 
> you may get. And you only get that when your cursor is far off the
> image center.

Yes, but zoom in-zoom out stability could be easily corrected with 
small mouse movement of the user (currently user has to do it anyway, 
see above). On the other hand with current behaviour you have to do 
huge mouse movement to make the scrolls.

You have to make the hoops _all the time_ to center area of interest 
(while zooming in point _not_ to area of interest, to push it). It is 
far from natural. So there are more costs than benefits currently.

And there is also one advantage in wished behaviour, well, by 
definition, you could scroll while zooming -- it is not great benefit 
for handicapped users, but no penalty though. For any other user it 
is great help because in one task you can perform two.

Sven, if you are not convinced, what about an option?  (yes, I am KDE 
user :-DDD):
[ ] scroll on zoom

* you could center the area of interest just by pointing at it and 
zooming in
* you would get context of the area just by pointing at it and zooming 

no scrolling, no artificial mouse movement. Just point and zoom.

Gimp-developer mailing list

Reply via email to