Re: [Usability] New Paradigm of Computing for GNOME 3.0





On 10 February 2010 15:38, Thorsten Wilms wrote:
On Wed, 2010-02-10 at 14:58 +0100, Diego Moya wrote:

>  - they still program using the interaction techniques of the late 70s
> (main application event loops, event-based triggering of subroutines,
> and throwing everything into a single application window with separate
> function points - all (mis)organized into lots of submenus.

I thought event loops belong to software architecture, not "interaction
techniques".
 
I should have said "event loops controlling event-based triggering of subroutines". Most desktop applications are defined by an event-based "collection of reactive commands activated throug menu entries, toolbars or keyboard shortcuts, with a collection of static widgets showing feedback", which is definitively an interaction technique.
 
Software architecture is definitely related to the interaction techniques supported - anything too complex will not get implemented using the wrong architecture. Complex interactions are best served by hierarchical state machines which allow creating and destroying interface elements on the fly. But most industrial-grade GUI tookits are based in the Widgets+events architecture, and GUI builders are little more than widget composers that help with the composition of static interfaces.
 
 
Anyway, what would be more modern approaches?
 
See Post-WIMP interfaces at Wikipedia (http://en.wikipedia.org/wiki/Post-WIMP), plus "Non Command User Interfaces" and "The humane interface" (http://en.wikipedia.org/wiki/The_Humane_Environment) for a good entry point. Physical interaction, zooming interfaces, continuous input and feedback, gestures recognition. 
 
Adobe-Macromedia Flash is a good and successful industrial example. Flash contains an advanced architecture for interaction, the Timeline. This makes coordinating animation through time much easier than with a reactive event loop plus timers.  Modern RAD web applications in Flash are ditching the timeline and using the classic widget approach, though. They call this progress.
 
 
 


> All private companies have that underlying goal. That is not
> incompatible with providing what the intended users need. I think
> Apple got it right in focusing on users with little computing
> experience/needs.

> The suggested solution (self-contained apps) is just one viable format
> for this, currently popular because of the success of the iPhone. As
> you say, this is not good for all users (only the majority of them) -
> so different solutions will evolve for the kind of users left behind.

I suspect that emphasizing applications has quite a lot to do with
pushing sell-able entities.
 
Maybe, maybe not. The Linux desktop emphasizes applications, even though they're Free and thus not sell-able. Maybe they are a relatively good and stable structure for grouping computing functions. This doesn't mean there can't be better ways.
 
 


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]