                           Changes in AIBB v2.0


Several changes have been made to this version of AIBB, and more are
forthcoming in the next version.  Note that versions 1.1-1.9 were never 
released (they had this annoying habit of displaying unwanted screen 
pyrotechnic effects...[read: crash and burn] :-) )

NOTICE:
I have an apology to make.  The previously released version of AIBB
(v1.0) has some inaccurate test comparison figures contained within it.
These inaccuracies are mainly in the area of the A3000, and can show it
to be a great deal slower than the A2500/30.  This is not correct.  In fact,
under most circumstances the A3000 should come up as slightly faster.
If this has in any way caused confusion I must apologize.  I have worked hard
to correct any of these inconsistancies, and hope that this version will
hold more to everyone's liking.  I don't like problems with my programs,
any more than people who may use them do, so if you find a problem, LET ME
KNOW!!!  I'll get looking into it right away.

OVERVIEW OF CHANGES:

    1.) Proper CPU identification in any circumstance.  No longer will the
        program give erroneous 68020 CPU IDs when in fact an 030 may be
        present.
    2.) The status of the Instruction and/or Data caches for your
        processor (provided you have an 020 or 030) is now displayed with
        the CPU type.  The ability to change the cache settings
        from within the program has also been added.
    3.) The ability to choose between comparisons by 68030 machines OR
        OTHER machine types has been added.  This allows for evaluations of
        the machine being tested over a broader range of available machines.
        Currently, the 'other' machine types are a 68020-based A2500/20 and
        a 68010-equipped Amiga 2000.
    4.) The addition of a memory alert.  If the program detects it has
        been loaded into CHIP or SLOW-FAST RAM, where bus contention may
        play havoc with the test results, it will inform you with a 
        small pop-up requester.  Note that although you will be able to 
        tell if the main program is in CHIP or SLOW-FAST RAM, you may
        not be able to tell if some of the tests are run in that medium.
        This is because some of the tests require allocation of memory on
        their own, for their own purposes.  The only way that this would
        happen, however, would be if the main program managed to use up
        most of the available FAST RAM, leaving little FAST type memory
        for the test's needs.  The easiest way to avoid any problems here
        is just to be sure you are not running the program on the last
        dregs of your FAST RAM.
    5.) There is an option now under the 'Test Options' menu which allows
        you to double the scale of the graph.  This is to allow people 
        lucky enough to have REALLY fast boards to fit themselves into
        the graph scale. (Read: If you get an 040 board sometime, it will
        probably fit on the scale...though you don't DESERVE to fit on it
        if you get one of those wonder beasts... :-) ).
    6.) A change has been made in the A2000 base results and how they are
        interpreted.  For full information, read the documentation file.
    7.) A log file saving option has been added for the saving of test
        results.  It requires the use of the 'req.library' or the 
        'arp.library'.  Please see the documentation for full details.
    8.) A small bug with the 'About...' requester has been (hopefully)
        corrected.  Some people were reporting that the requester wouldn't
        let them select it's 'Continue' gadget.  I believe I've corrected
        that problem in this version. (I was running into timing problems
        with Intuition)

PLANNED FOR THE NEXT VERSION:
    The next version of AIBB will probably be 3.0, and will basically be
    a revamp of this version, but with full internal figures for the
    comparison machines running AmigaOS 2.0.  The delay is due to the
    fact the CBM hasn't yet burned 2.0 in ROM (as of this writing), and
    I want to give it time to get around first.  There will be a difference
    in the comparisons, as 2.0's ROM functions are in some areas faster than
    the ones in 1.3.  [And oddly, some tests are running slower in preliminary
    examinations.  There may be a few interim changes in AIBB (2.x, etc..)
    depending on if any bugs turn up (gasp..no... :-) ).  I'm intending
    AIBB 2.0 to not just be 2.0-specific either.  It will adjust itself 
    depending on the OS you are running. 

--LaMonte
