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