#8482 NORM 9.1.0: When rotating screen orientation, cursor movement should remain relative to the touchpad. (was: When rotating screen orientation, it would be more helpful if cursor movement stayed un-rotated.)

[Zarro Boogs per Child]

#8482: When rotating screen orientation, cursor movement should remain relative to
the touchpad.
---------------------------+------------------------------------------------
   Reporter:  garycmartin  |       Owner:  marco                            
       Type:  enhancement  |      Status:  new                              
   Priority:  normal       |   Milestone:  9.1.0                            
  Component:  sugar        |     Version:  Development build as of this date
 Resolution:               |    Keywords:                                   
Next_action:  code         |    Verified:  0                                
  Blockedby:               |    Blocking:                                   
---------------------------+------------------------------------------------
Changes (by Eben):

  * next_action:  communicate => code
 * cc: Eben (added)
  * component:  interface-design => sugar
  * milestone:  Not Triaged => 9.1.0
  * owner:  Eben => marco


Comment:

 Thanks for the thorough enumeration.  Let me take a stab at reigning
 things into a reasonable set; as many potential configurations as there
 are, not all seem usable (or desirable).

 As you mention, we have hardware limitations in detectable orientations.
 I believe we know (with boolean resolution) whether or not the laptop is
 in handheld mode.  This eliminates the possibility of handling the "almost
 in tablet mode" case.  I find this acceptable, because I don't think such
 a configuration serves any reasonable use cases; it sounds instead like an
 attempt to use an activity in handheld mode which doesn't support it, and
 therefore doesn't take advantage of any of the conveniences that handheld
 mode is designed for.  On the other hand, what we do know with certainty
 is the software orientation of the screen.

 Here's my revised enumeration of the possible physical orientations of the
 device:

  1. Laptop mode
  2. Handheld mode
  3. Hybrid mode (book mode, portrait mode?)

 Laptop mode is the conventional orientation, to be used with an open
 laptop positioned on a table or lap.  Screen rotation doesn't make sense
 here, so neither does transformation of input. Handheld mode is the next
 most common orientation, in which the physical display is rotated 180º and
 closed flat.  Any screen rotation makes sense here, but the touchpad is
 inaccessible, so the point is again moot.  Finally, a potentially usable
 scenario which hadn't been previously considered was handheld use of the
 laptop in, more or less, laptop mode.  (That is, pick up the laptop as you
 might a book, with with the screen in one hand and the base in the other.)
 This can be done with a physical rotation of 90º or -90º, and might depend
 on hand preference.

 This last mode is interesting, since, while it makes the keyboard mostly
 unusable, it does position the thumb of one hand directly over the
 touchpad and mouse buttons.  Unfortunately, we can't ''detect'' this mode,
 since we don't have an orientation sensor, but it could still have useful
 applications.  The interesting thing to note (as you did), is that we
 desire is to retain, regardless of the screen rotation, a relative
 physical mapping to the screen, which therefore requires inverse rotation
 of the input device with respect to the screen rotation.

 Incidentally, one modification to this one basic rule will cover all of
 your peculiar physical configurations: simply rotate in the same direction
 as the screen rotation, instead of the inverse direction (since we rotate
 the screen 180º). I still don't necessarily find them useful, and my aim
 here is to narrow down the cases we need to talk/think  about to the
 minimum, so I'd recommend sticking to the simple inverse rule alone.

 Note that the initial subject actually misspoke; the desire is to rotate
 the touchpad input so as to keep the relative mapping to the screen; It's
 presently unrotated (up remains up, but in the new screen orientation).

-- 
Ticket URL: <http://dev.laptop.org/ticket/8482#comment:3>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system