Usability suggestions about the new authorization API for KDE

Hello!

I'm the Google Summer of Code student who is writing the new  
authorization library for kde. It currently is in the subversion  
repository at playground/base/libkauth.
I'm writing to you because the library is almost complete and I'm  
going to start the step 2 of my project: addition to the GUI elements.

In detail, my library allows developers to easily write an application  
that require high privileges (for example kcm modules, partition  
editors, package manager front-ends, etc..), by splitting the code  
into a small isolated helper tool that runs as root, and the rest of  
the code (with the GUI and other things) that runs as normal user. The  
authorization and user authentication is performed in a cross-platform  
way, using policykit on linux and native APIs on other platforms.
This adds more security to the system, since the GUI code doesn't run  
as root anymore.

What I have to do now, is to add some new behaviours to GUI components  
like KAction, KPushButton, etc... to let the developer to instantly  
begin to use the new API.

The feedback that these components should send to the user include,  
but are not limited to:
- That the button they're clicking will trigger some high-privileged  
task (possibly requiring authentication)
- Whether the user is currently authorized to do the action or not.

What I ask you are suggestions about how these visual elements shoud  
look and how they should behave.

I have some idea. The first, for a KPushButton, is to add an icon  
(something like a lock or a key) to represent that the push button wil  
trigger an authentication. This would be useful for actions like Save  
on a kcm module.
But another pattern is possible. As in (for example) preference panes  
in Mac OS X, an unauthorized user would find all the controls in the  
window disabled until it unlocks a toggle button (which will have as  
the icon a close or open lock). Unlocking will trigger the  
authentication which, if successful, will enable the controls in the  
window.

Which of these behaviours is better from a usability point of view? Of  
course, both could be implemented to be choosen by the developer.

Thank you,

Nicola
_______________________________________________
kde-usability mailing list
kde-usability@xxxxxxx
https://mail.kde.org/mailman/listinfo/kde-usability