Hi Lukas: There are 4 tests that were failing in Windows and I debugged them. The tests are the 4 tests of the class OBKeyBindingsTest

The problem in windows is this. In your tests, you do:

modifier: mod keycode: code
�� ^ KeyboardEvent new
�� setType: 'keystroke'
�� buttons: mod
�� position: 123 @ 456
�� keyValue: code
�� charCode: code
�� hand: nil
�� stamp: nil

As you can see, you don't set a codeScan (it is nil).

In Paragraph >> dispatchOnKeyEvent: keyEvent with: typeAheadStream��

You do "OSPlatform current virtualKey: keyEvent scanCode."

In mac, everything is ok as it uses the method�

OSPlatform >> virtualKey: virtualKeyCode
�� "Subclass responsibility to override if necessary"
�� ^nil

But in Windows, it uses:

virtualKey: virtualKeyCode
�� "Win32Platform virtualKey: $C charCode"

�� (virtualKeyCode <=� 90 "$Z charCode"
�� and: [virtualKeyCode >=� 65 "$A "])
�� ifFalse: [^nil].
�� "#($a $b $c $d $e $f $g $h $i $j $k $l $m $n $o $p $q $r $s $t $u $v $w $x $y $z)"
��
�� ^(#($a nil $c $d nil $f $g nil nil nil nil $l $m $n nil $p nil nil $s nil nil $v nil $x nil $z)
�� at: virtualKeyCode-64) ifNotNil: [:char | char charCode]

There, is a DNU as nil (virtualKeyCode)�� dnu:� <=

So, the solution I think is to set a scanCode in the keyEvents you generate in the tests. I don't know which number is correct for each type of event, but I tried just with 0 and tests are green.

modifier: mod keycode: code
�� ^ KeyboardEvent new
�� setType: 'keystroke'
�� buttons: mod
�� position: 123 @ 456
�� keyValue: code
�� charCode: code
�� hand: nil
�� stamp: nil;
scanCode: 0

What do you think? is there a better fix ?

Cheers

Mariano