Fresh Fish 8/FreshFishVol8-CD1.bin/useful/reviews/hardware/emulation/emplant_3

Internet Message Format | 1993-06-25 | 8KB

Path: menudo.uh.edu!usenet
From: [e-mail] (John Harris)
Newsgroups: comp.sys.amiga.reviews
Subject: Re: REVIEW: Emplant (and comparison with AMax II)
Followup-To: comp.sys.amiga.emulations
Date: 25 Jun 1993 19:59:36 GMT
Organization: The Amiga Online Review Column - ed. Daniel Barrett
Lines: 143
Sender: [e-mail] (comp.sys.amiga.reviews moderator)
Distribution: world
Message-ID: <20flf8$[e-mail]>
Reply-To: [e-mail] (John Harris)
NNTP-Posting-Host: karazm.math.uh.edu
Keywords: hardware, emulator, Macintosh, commercial, followup


    This is an update to my EMPLANT review posted in c.s.a.reviews on
April 4, 1993.  I have been waiting for a particular enhancement to the
product -- that of making it 32-bit clean -- and have wanted to hold off a
formal update until that is available.  Since it is taking longer than
expected, and several other changes have taken place, I thought it best to
write a quick update now.  A more thorough update will be posted when EMPLANT
achieves 32-bit clean status.

    This article covers EMPLANT version 2.91.

    [MODERATOR'S NOTE:  John's original review is in the c.s.a.reviews
    archives on math.uh.edu in the file named Emplant_2.  - Dan]


OMISSIONS IN THE ORIGINAL REVIEW

    I should have mentioned the cost of updating EMPLANT.  With almost
30 software revisions in the 8 month life of EMPLANT, it is important to
note that all of these updates have been provided free of charge over the
on-line services.  Plus one disk, about version 1.5 or so, was mailed out
for free as well.  My apologies to those readers who wondered how much money
it was going to cost to stay current with EMPLANT software updates!  And
thank you for bringing the omission to my attention.


IMPROVEMENTS

    Many improvements have been made to the Mac II emulation software.
Here are some of the highlights.

    EMPLANT screen modes are now available in overscan.

    There is a video driver for the Retina board that provides 256-color
and full 24-bit support.  I got to test this board very briefly and found
the 24-bit mode very slow.  However, the 256-color mode is nice and quick.
It's much faster than 256-colors on an AGA machine, and even faster than the
16-color mode on my native ECS chipset.  The driver is supposed to get
several times faster in the near future.  I was also able to run PhotoShop
in both 256-color and 24-bit mode; and although I didn't run extensive
tests, I found no problems.  I will write more on this in a future update.
Video drivers for other boards are also due shortly.

    Sybil now writes Mac format disks.  There are also new routines in
place that do not require the calibration step.  Personally, I have had more
trouble with the new routines than I had with the old ones.  Once I had the
old setup calibrated, everything worked fine.  Now, if the software has a
bit of trouble reading a weak spot on a Mac disk, it tries to auto-calibrate
which throws it away from the correct setting, causing a permanent failure
of the copy process.  Before, it would retry the same place a bit, and could
ultimately get past it.  Most of the disks I have tried to read recently
have given me problems, and I'm still waiting to get some questions answered
by UU.  They are supposed to make another software enhancement to get past
hard-to-read sections on a disk, and this may fix the problems I'm having.

    The disk convert software can read disks from a real Mac drive
connected with the AMax cartridge, Mac-2-Dos, or the public domain hack.
This ability is still not available within the emulation, but it should be
available soon.

    The error detection and reporting of problems related to an
individual's EMPLANT setup have been greatly improved.  There is also a hard
drive installation program provided, so getting a working setup is much
easier.

BUG FIXES

    All of the 32-bit clean and MMU problems are still here.  However,
everything else listed in the 'PROBLEMS' section of my original review have
been fixed, or at least are being addressed.

    The real time clock is fixed.

    The floppy drives are no longer locked out when running the
emulation.  They can be switched from Amiga mode to Mac mode while the
emulation is running.

    The Sybil routines can be switched off when not in use, which
eliminates problems with unwanted printing and paper ejection.

    The VBR (interrupt vector table that the CPU uses) is now separate
between Mac and Amiga applications.  It has also been moved out of its
native location at address $0, and moved into Fast RAM.  This has some
plusses and minuses.  On the plus side, everything in the emulation is
faster, since the VBR is now in Fast RAM instead of Chip RAM.  Also, there
were some Mac programs, mostly games, that stuffed values directly into the
VBR which crashed the Amiga system.  Now with separated VBR's, things
co-exist more peacefully.  The only drawback is that some Amiga programs --
mostly games -- rely on having the VBR at location 0!  U.U. has provided a
program to move the VBR back to 0, but it's a bit of an annoyance to have to
remember to run it.

    EMPLANT now traps out all of the processor exceptions.  This has
made any program crashes I have experienced exit in a graceful way.  Where
the machine used to just lock up or reset, it now presents a Mac 'bomb'
dialog with an option to restart.  The Amiga side is still active here, so
it's a good idea to save anything that's in progress!  Unfortunately, I
can't run this new version, since the TickerWatcher program that I have to
use everyday bombs with a 'Bus Error'.  This is the same program that gave me
problems in earlier EMPLANT versions, and now it's back to not working
again.  I guess this is a good time to restate that EMPLANT is still
undergoing a lot of changes; and until things are in a more finished state,
quirks like this are inevitable.

FUTURE ANNOUNCEMENTS

    A new and very complete manual is reportedly being printed, and will
be mailed free of charge to all registered users.

    U.U. has announced that emulation modules for Commodore 64 and Atari
8-bit computers will be provided at *no charge* when they become available.

    U.U. has formally announced how they will provide IBM emulation
capability.  They are planning to do a complete cross compile of the entire
program's worth of 80x86 code into 680x0 code at program load time.  If this
works, IBM programs would run very quickly compared to other software
emulators, and theoretically, even faster than hardware emulators if you
have a fast Amiga.  The hardware emulators are only 16-bit machines.  While
I think this approach is possible in a broad sense, I have doubts about
overall compatibility.  Jim Drew claims to be able to handle self-modifying
code, self-unpacking programs, and other tricks, but only time will tell if
this will really be possible.  The way I look at it, most programs are
written in high-level languages these days and probably have few tricks that
would cause the emulation problems.  I also suspect that any non-working
programs, if popular enough, could have support for them individually
patched in.  As usual, expect games to have the toughest time in this
emulation environment.


    Thank you for all of the wonderful feedback, that I have received
regarding my EMPLANT review.  If you have any more questions, please feel
free to contact me.

John Harris
[e-mail]
GEnie: J.HARRIS32

---

   Daniel Barrett, Moderator, comp.sys.amiga.reviews
   Send reviews to:    [e-mail]
   Request information:    [e-mail]
   Moderator mail:    [e-mail]