[Bf-funboard] Re: Toolbox Design 3

[Thorsten Wilms]

On Thu, Oct 09, 2003 at 05:37:19PM +0200, Guillermo S. Romero / Familia Romero wrote:
> t_w_@freenet.de (2003-10-08 at 2150.59 +0200):
> > Hi!
> > 
> > An update to be my Toolbox work:
> > 
> > http://wrstud.urz.uni-wuppertal.de/~ka0394/en/blender_ui/
> > 
> > Please give me some feedback!
> 
> Design 1: crowded and forcing different scanning procedures. It is
> like a pie menu to start, and later worse than normal menus. Reducing
> time by less distance is fine, but you lose time cos the things are
> crowded and the order looks random at first.
> 
> Design 2: it is like pie menus but the previous entry is lost, right?
> How do you go back or know where are you? And again, pie menu for
> first step, then back to normal but crowded in some cases.

Yes, first entry is lost. To go back leave menu and hit spacebar again. 
Not nice for learning the contents, but shouldn't be a problem after some 
getting used to.
 
> Design 3: crowded, but a bit less too. From where did you got the idea
> that a two column list is easy? Or that you have to split things in
> two groups?

I wanted to do something against the drawbacks of a linear menu.
For up and down directions selection happens now in 2 parts:
left/right and the line (height). It might be harder to scan, 
but will be more effective. 

> In all cases: the command line would collide with hotchars function if
> it gets added to all menus, and you removed the shortcut feedback.

It's only meant for the toolbox, and there accesible only at the toplevel.
For the cross that would mean no "hotchar" access to the four toplevels. 
I you want keyboard navigation, no prblem: cursor-keys (maybe numpad for 
more than 4 way menus, if we need some).

> 
> What I know for sure about menus:
> 
> Pie menus are based in navigating by direction, and possibly giving
> options to go back. Distance traveled is not important at all.
> 
> Normal menus are based in single lists in reading order (left to right
> and top to bottom for languages that are read that way), ordered by
> similar functions and using separators or blanks to add clarity.
> 
> In both cases, they do not mix concepts, which you do. Your idea
> sounds like hybrid, and dunno if this one of the cases were hybrid is
> good at the end.
> 
> So I agree with the other post: where can this be tested beyond
> mockups? The images alone do not convince me. I want to test how such
> lack of white space or separators works, and how mixing two methods
> works.

Like with other things new this can't be tested at the moment.
You need to have some imagination. I'm not a develper, so I can't 
offer a prototype.

I just wan't to provide alternative ideas. Going straight with what is 
already known will cut out the possibility to find better ways.


---
Thorsten