------------------------------------------------------------------------------
 OD-90228                         SUGGESTION                       "PRIDE"-IF
------------------------------------------------------------------------------
 JUN 16, 1997                   RELEASE  1.2.0                        PAGE  1

       ****************************************************************
       *               MI-10061 ONE KEYBOARD PER KERNEL               *
       ****************************************************************

 REQUIRED DATE- ASAP          USER PRIORITY WEIGHT-  26    RANKING-  0009
 DATE RECEIVED- JUN 02, 1997  REQUEST STANDING  - REVIEW   AS OF- JUN 16, 1997
                              TYPE- MOD/IMP

 DESCRIPTION-

        Do other multitasking operating systems assume the user has any
        use for maintaining separate shift states for each window/
        application?  I would be greatly surprised.  I can understand
        that some users might have a use for the result of that
        assumption.  I cannot understand why IBM thinks all OS/2 users
        want their system to boot contrary to the BIOS setting for their
        system - e.g. NUMLOCK set ON.  It is past time to do at least
        one of two things to all versions of OS/2:

        1 - Change Warp to assume the user wants the boot up shift
            states to be maintained, and/or;

        2 - Add two new optional CONFIG.SYS set statements to allow the
            user to define shift states:

            A:  keyboard#=[system,sessions] sets system's assumption of
                whether shift states will be separately maintained for
                each session (keyboard#=sessions)= or will remain
                unchanged from session to session (keyboard#=system,
                the new default)

            B:  shiftstates=[on,off],[on,off],[on,off] sets boot up
                initialization states for CAPSLOCK,NUMLOCK,SCROLLOCK,
                and if keyboard#=sessions, the initial settings
                for each new session.  These changes will enable OS/2
                users to efficiently use their keyboards, instead of
                one of two things that now happens:  either bogging
                down the system with a daemon utility to fight OS/2's
                insistent flipping shift states to off; or correcting
                keystroke errors caused by different shift states in
                different sessions.

        ATTRIBUTES:

        MEDIUM Value
        Desired ASAP

        Submitted by:

        Vern Tonnesen
        Weeki Wachee, FL, USA
        jd.cpa@fibernet.com

                           * * * END OF OUTPUT * * *