[Dapper] Mouse

[Alan McKinnon]

On Friday 17 March 2006 06:17, Peter Garrett wrote:
> On Thu, 16 Mar 2006 14:01:24 +0200
> Alan McKinnon <alan at linuxholdings.co.za> wrote:
>
> [ And Peter Garrett previously wrote ]
>
> > > I take your point about mouse settings  - but my point remains
> > > that _for a naive user_ , clicking behaviour would be expected
> > > to be under a "mouse" heading rather than a "file management"
> > > setting.
> >
> > Without meaning to throw a cat in amongst the pigeons, I'd
> > suggest that click behaviour goes in a third category, Desktop.
> > To illustrate:
> >
> > Desktop:
> > 	Click behaviour
> > 	Menus
> > 	Background
> > 	etc
>
> < snipped>
>
> Yes, that makes sense - in fact if it were labelled "Click
> Behaviour" that would be pretty clear - *but*  there are, in fact,
> "Click Behaviour" settings already under "Mouse": -  just not
> single/double click settings.
>
> Really, I feel it's just a question of making this discoverable to
> the new user - I'm not looking at it from a "logical" viewpoint so
> much as a "where can a user find the setting?" viewpoint. The new
> user cares little where it is, just as long as it is obvious!

True, the whole point of an interface is to make it possible to get 
something done, and you need a usability study to know how users 
react to it. My own experiences show me that user think click 
behaviour belongs in a category for the desktop (or gui) as a whole - 
but I can;t be sure that still holds true for the majority of users.

> There may be good technical reasons why the "Mouse" dialogue can't
> handle this function as well. I'm not a coder so I can't say. It
> does seem to me that either having the double/single click option
> in both places ( programmers don't like redundancy, but users might
> in a case like this), or at least a pointer to the place to do it,
> might be an idea. For example: a button to click that launches that
> function from within the "mouse" dialogue.

Implementing the best solution discovered by survey is the smallest 
problem. Even if it does mean ripping out chunks of architecture code 
and replacing it with something else. Coders do this all the time, 
and even if they complain bitterly, they go along with the change if 
it has provable benefits.

The trick is that a change like this has to be implemented upstream 
and that involves proving to the gnome developers that you have a 
valid point. Which brings us back to a usability study.

-- 
Alan McKinnon
alan at linuxholdings dot co dot za
+27 82, double three seven, one nine three five