From kde-devel  Tue Jul 13 08:48:18 1999
From: "Markus Holzem" <markus () holzem ! de>
Date: Tue, 13 Jul 1999 08:48:18 +0000
To: kde-devel
Subject: FYI: Why selecting is everywhere
X-MARC-Message: https://marc.info/?l=kde-devel&m=93185561314280

First I want to apologize for my outburst but I was not aware that
only few developers are on kde-look.

It is difficult to transport, what the problem with single click (SC)
and double click (DC) of the left mouse button (LMB) is. Oops - lamps
burning? Again a stupid mail on this matter? Please make the effort to
understand it. I don't have the possibility to show you video tapes of
users (novice, advanced, expert) sitting in front of KDE to proof the
difficulties they have, but perhaps sensible arguments can help too:

Whenever the user is confronted simultaneously with more than one item
he needs an option to

 * select one item (usually done with SC)
 * select successive items (usually done with SHIFT + SC)
 * select non-successive items (usually done with CTRL + SC)

Examples:

 * user searches for specific files and gets back a list. He selects
   one or more files and e.g. moves/copies/deletes them.
 * user receives mails and wants to organize a few in a subfolder.
   He selects them and puts them to the destination
 * user draws with an OO graphic application. He selects some objects
   and recolours them.
 * user wants to change a specific item in a dialog (e.g. list or
   combo box) he selects the dialog item and changes the data.

These are only a few examples I can think of. All selections are
performed by a LMB SC.

Whenever a user wants to perform an action on the selection he selects
it from a toolbar or a menu bar. But some actions are that usual, that
it was decided, that the user should have a simple shortcut to perform
the action immediately on the selection:

 * a DC on the selection performs what is called the default action.
   For a file manager it would be to open the file, for a dialog box it
   is usually to close the dialog box with the default button (i.e. OK),
   for a tree node to open the sub tree, ...
 * a context menu that is popped up, when the user presses the right
   mouse button (RMB).

That gives the user the following impression of the user interface:

 * LMB SC: select items (modifiable by SHIFT and CTRL)
 * LMB DC: perform some very common action on the selected items
 * RMB SC: pop up a context menu that shows the most important
           actions that can be performed on the selection (and
           perhaps the default action in bold face).

This scheme is kind of "burnt in". After some time when the novice
progresses and becomes an advanced user it even works - at least
partially - on a subconscious level and he uses this scheme on all
applications without thinking.

For the novice it might be even more important, because it means
that he cannot destroy something by simply clicking on it. He becomes
more bold to try out new things.

That is basic HCI (human computer interaction): Learn once - use
everywhere. The user should not need to have the concept of different
applications (and he often might not get the point if he was able).
He learns how a desktop environment (DE) works once and can apply
it on all other applications of the same DE.

This is were user interface guidelines come in. They fix, how the
user is supposed to work with a DE. And guidelines don't have the
concept of different applications either. This is the reason why
I insist that much, that from the users perspective the behaviour
of the desktop and kfm are so problematic. Sometimes the LMB SC
performs an action and the user has no concept why this is so.

Everyone talks nowadays about "intuitive" usage of user interfaces.
Most HCI designers are happy when they get to the point when it
becomes memorable. And a DE becomes more memorable when the mouse
does not behave differently in each application. It part of the
term "Look & Feel": Feel.

Just to clarify the matter of SC vs. DC.

Markus