18/11/2003
|
Finally, I got a ARM OS5 Device (Palm Tungsen|E). I'm still planning time to upgrade the Emulator for a Armlet Version. (Hilary)
|
09/4/2003
|
I added a CAPS LOCK/UNLOCK function. It is independent of the system caps setting.
|
09/3/2003
|
I bumped the minimum OS version to 3.5. Anything lower rejects the call to
allocate 64K of memory used as the Apple's main memory bank.
Fixed a few menu bugs and JogDial activation with the dialog boxes open.
|
08/29/2003
|
Cleaned up the sound a little. Along with a slight change in
the refresh rate implementation, it sounds reasonable (for a
PalmPilot, anyway). The refresh rate increase/decrease inversely
affects interactivity. Your emulated apps may run a little slower
but grafitti will be recognized better. Especially useful for you
OS5 people. If you have a fast OS3/OS4 device, then you can speed up
your apps by decreasing refresh rate. Interactivity shouldn't suffer
for modest refresh rate decrements.
|
08/27/2003
|
I've wrapped a few user requests into a binary release.
I've increased the emulated cycles between event checking
so that 33 MHz devices shouldn't need overclocking for
decent performance. Interactivity doesn't seem to suffer.
I have also made a request to developers for help in adding
an Appalm Manager application to enhance the features of Appalm.
|
08/16/2003
|
I think I've got the LotA bug resolved. In
the process, I overwrote the wrong disk as the character
disk. So, I've added an explicit WriteEnable checkbox
to the disk load dialog. The WriteEnable will not
override the read-only flag on the database.
|
08/14/2003
|
It is now official - Appalm ][ is out of
code space. The straw that broke the camel's
back was JogDial support. It is pretty cool, though.
The JogDial emulates Paddle 0. Press the JogDial for
button press. If you have a back button, that will
toggle between Paddle 0 and Paddle 1. I know it is
bad form to make the back button do something non-standard,
but it is too easy to press by accident and
drop you back in the launcher just as you were about to
annihilate that last alien. Works great with games like
Space Invaders, Galaxians, Sneakers and even Little Brick Out :-)
|
08/14/2003
|
Sound does work on the Clie. I accidently
turned of the sound output in the 68VZ328
Peripheral Control Register thinking that
SndDoCmd used the PWM2 circuit, not the
PWM1 circuit. Oddly, a hard reset didn't
change it back, I had to rewrite the default
value back by hand. Anyway, the Clie is buzzing
away. The sound implementation still sounds bad
though.
|
08/12/2003
|
I broke greyscale recently and just fixed it.
To make up for my boneheadedness, I re-implemented
greyscale in assembly. Much faster (and it works).
|
08/11/2003
|
I ported the hires device color
code to standard res device support
for text and hires modes. All of
you with Palm Vs can now enjoy your
games in color instead of
greyscale. Should be as fast as the
hires device code.
|
08/9/2003
|
More works than I thought. Thanks
to a tip from mmulhern about
copy-protected disks, Wizardry and
the Ultimas work great. You will
need to set the read-only property
on the boot disk image databases to
get them to work. I will update the
working list. Now, if I could get
sound to work on the Clie we would
be much closer to a 1.0 release...
|
08/6/2003
|
OS5 High Resolution should be
fixed. But We don't have any OS5
Device. Anyone wanna to test. Pls
send email to us. PLS HELP!
|
08/1/2003
|
Fixed the really annoying bug with
drive selection. Interestingly, the
addition of the two drive controls
helped debug the problem with the
wrong drive light coming on. Also
improved the text performance a
little (and reduced the code size
somewhat).
|
07/29/2003
|
80 column card support for HiRes
devices (no double hires yet).
Attempt at OS5 high density support
and a fix for dsk2pdb for Borland
binary (I hope).
|
07/12/2003
|
Initial sound support using
SndDoCmd. It doesn't work on my
Clie (quiet ticks only) but it does
on my IIIxe. Since it is quite a
CPU hog, turn it off by setting
game volume to off in system
preferences.
|
07/11/2003
|
A language card fix that cleans up
SuperZaxxon and gets SkyFox
working. I thought SkyFox was the
coolest game when it came out, but
I could never get the hang of
playing it. I find it much easier
to play on the Clie with the game
pad (and still quite cool). Ultima
IV runs better too. Many games
require the extended 80 column card
on the //e for the double hires
mode. Kings Quest and Might &
Magic are the two that I looked at.
Once I get Wizardry working....
|
07/10/2003
|
Performance tweaking and graphics
mode support improvements. I added
lo-res graphics mode for The HiRes
Color devices and switched the
LoRes Mono devices to 2BPP mode.
Much improved quality. When
performance tuning, I uncovered a
not-so-nice feature of the
peripheral device drivers. Those
PRC files that make your external
keyboard work transparently with
the other apps. Turns out, they
suck a lot of CPU power, even when
they are disabled. The thumb
keyboard is the worst - 50%
performance degradation with it
resident. The folding keyboard and
the game controller suck up about
10% each. Amazing. What the heck
are they doing? Remove those files
if performance is important.
|
06/24/2003
|
Whew. Ultima IV is the ultimate
pain when it comes to language card
access. it alternates between bank
1 and bank 2 for every image byte
access. So, after implementing a
crazy scheme to update just the
page table entries for R/W and
insert special versions of the
opcodes that could change the
program counter into language card
space, then fix everything up when
that happens and make it fit on a
Palm and debug it all, it now works
well enough to play. Apple Pascal
now boots too :-) And all because
Hilary wanted to play Ultima IV. I
also added a little tweak here and
there to make interactivity a
little better. Karateka kicks @$$
now.
|
06/22/2003
|
Dual disks implemented. There seems
to be an issue where you
sometimes have to mount the
disk twice to get it to read. Work
in progress...
|
06/20/2003
|
Disk writes are working now. The
disk image data base ReadOnly
attribute controls the
write-protect. Pretty cool. Maybe
I'll add support for dual disks.
|
06/20/2003
|
Wow, the latest CVS can boot ProDOS
at Sony Clie T650 99Mhz with
9-10seconds. Amazing. the Speed is
nearly same as a REAL
APPLE][. But it has bugs.
Please stay tune.
|
06/19/2003
|
Big check-in. Disk I/O is much
faster. My first attempt to emulate
the disk rotation has met with
better than expected results. I
also found my problem with Lode
Runner - I was using non-strict
stack emulation and paid the price.
Luckily I anticipated this with a
#define that switches between my
strict and non-strict
implementations. We now are using
strict stack manipulation, but I
didn't see much speed difference
anyway.
|
06/18/2003
|
Planning on major disk I/O rewrite.
I have changed the raw track size.
This breaks current DSK.PDB files
so you will need to re-create them.
Hopefully you will see a big
speedup shortly.
|
06/17/2003
|
All new
implementation. Most PalmOS
hardware should be able to run the
emulator. Many outstanding bugs
exist. Performance is close
(slightly faster ?) to native Apple
//e on 66 MHz Clie with HiRes
color. Still only working with raw,
nibblized disk format. Command line
tool for converting DSK ->
PDB.
- ProDos boots very fast (but
broken 80 column probe)
- Choplifter runs great
- Karateka runs great
- Sammy Lightfoot runs great
- Lode Runner missing opcodes
:-(
- Wizardry = Crash :-(
|
02/05/2002
|
I am planning to make a new release
on 30/05/2002. It will including a
ROM/Disk Converter (Currently only
Support DSK format, no Nibble)
Tools for Win32. Stay Tune.
|
29/04/2002
|
I have fixed some bugs that allow
to switch back to Other Palm
Applications. Also, no Fatal
Exception after pressing Reset.
|
10/04/2002
|
Since the Speed of Emulator is not
Enough to Play any Apple][ Game. So
We will try to use M68000 Register
to Store the Value of 6502
Register. It may improve the speed
of the Emulator. Hope that It can
play a few games after the
modification.
|
06/04/2002
|
We have started to
implement Hi-Res Drawing Function.
PalmApple can successfully
boot
- ProDos
- Ultima IV
- Karateka
B ut Most of
it needs about 20 minutes to boot
in POSE. So More Performance
Enchancing is needed so that games
are playable.
|
20/03/2002
|
After Fixing a few bugs, Currently,
Palm IIIc with 37Mhz (Overclock
with AfterBurner) can boot First
ProDos Splash Screen about 35
seconds. It is GREAT!. It is much
FASTER than Previous Version.
Previous version needs about 2
minutes to boot!
|
03/04/2002
|
Fixing lots of Bugs, including 16K
RAM Card + Video Soft Switch. Since
ProDos will test PAGE2 Video
Memory. I have implemented it by
filling a non-readable &
non-writable memory for $400
Memory. For the Right Now, ProDos
can BOOT without any problem!! It
can perform CAT & CATALOG
command! YEAH!
I am preparing for the new release
and upload to sourceforge +
documentation that allows other
people to join the projects.
|
20/03/2002
|
After Fixing a few bugs, Currently,
Palm IIIc with 37Mhz (Overclock
with AfterBurner) can boot First
ProDos Splash Screen about 35
seconds. It is GREAT!. It is much
FASTER than Previous Version.
Previous version needs about 2
minutes to boot!
|
19/03/2002
|
One of the emulator bug make things
happen very differently. ProDos can
boot in the Sony Clie ROM but
cannot boot in PalmIIIc ROM. It is
strange! So every assembly code you
have written should be checked very
carefully.
Currenly, I need to implement
Enhanced IIe Feature to expand more
64K memory. Since ProDos needs
Enhanced IIe Features.
|
18/03/2002
|
A Lot of Bug Fix in the 6502
Emulation Core. We should be very
careful when writing assembly
language. since the compiler can't
help you to check much of your
code.
|
06/03/2002
|
We can use it by C Pointer. Use
MemSemaphoreReserve() to enable
direct access the heap storage. :p
But it is very dangerous. It can
overwrite everything in the Palm.
|
05/03/2002
|
I have found FtrPtrNew to Create a
64K Memory Block. But it is not
accessible by using C Pointer. We
need to use DmWrite to access it.
|
19/02/2002
|
I have start to re-write all the
stuff of PalmApple That is using
inline Assembly to Implements it.
Also, DragonBall is M68000 Based
CPU. It got no CPU32 Instructions!
-__-!
|