Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Any pointers on emulating an Apple IIe (enhanced) in Mame on Mac OS X?

830 views
Skip to first unread message

Zellyn

unread,
Sep 11, 2016, 9:13:00 PM9/11/16
to
Hi folks,

I used homebrew to install homebrew/games/mame on my machine, which appears to be 0.177, and ran opensnoop to get lists of ROMs needed until it was happy and will run an Apple IIe.

However, it appears that it puts the keyboard into some raw SDL mode, and I can't figure out how to bring up the MAME/MESS UI. I believe on Windows, it would be "scrl lock" or "ins", but neither of those keys exist on my MacBook keyboard.

Has anyone gotten this working before?

Alternatively, does anyone know of a Mac OS emulator that does a really good IIe (enhanced) emulation? I am looking to add IIe / IIe Enhanced support to OpenEmulator, but I need a relatively high-fidelity "ground truth" to test against, since I don't have an actual IIe.

Thanks for your help,

Zellyn

Richard Thiebaud

unread,
Sep 11, 2016, 9:18:13 PM9/11/16
to
I've very successfully under LInux run Applewin under WINE. It seems
likely that this would work under MacOS. Just a thought.

Zellyn

unread,
Sep 11, 2016, 9:21:46 PM9/11/16
to
On Sunday, September 11, 2016 at 9:18:13 PM UTC-4, Richard Thiebaud wrote:
> I've very successfully under LInux run Applewin under WINE. It seems
> likely that this would work under MacOS. Just a thought.

Oh, it's running quite successfully… I can even play Ancient Legends like this: mame -rompath ~/mame/roms/ apple2ee -flop1 ~/Documents/a2-disks/Ancient\ Legends\ -\ Disk\ 1.dsk -flop2 ~/Documents/a2-disks/Ancient\ Legends\ -\ Disk\ 2.dsk

So I guess as long as I don't want to dynamically switch disks around or change emulation options or anything, it'll work fine.

Zellyn

Richard Thiebaud

unread,
Sep 11, 2016, 10:04:44 PM9/11/16
to
On 9/11/2016 9:12 PM, Zellyn wrote:
Virtual ][ (www.virtualii.com/) looks good but it is not free. A full
license is $40.

Peter Ferrie

unread,
Sep 11, 2016, 10:37:30 PM9/11/16
to
mame.ini:
uimodekey SCRLOCK

change it to a key that you have but don't use much (e.g. LEFTBRACKET), and you'll be fine.

Zellyn

unread,
Sep 12, 2016, 9:44:50 AM9/12/16
to
Hmmm. That doesn't seem to work either… :-( I can run mame and see that the config change took effect, but it doesn't work while running the emulator.

sicklittlemonkey

unread,
Sep 12, 2016, 7:15:07 PM9/12/16
to
On Monday, 12 September 2016 13:18:13 UTC+12, Richard Thiebaud wrote:
> I've very successfully under LInux run Applewin under WINE. It seems
> likely that this would work under MacOS. Just a thought.

It's really easy to get AppleWin working on MacOS:
http://www.appleoldies.ca/Mac/AppleWin.zip

That version is a couple of years old, but you can copy the latest AppleWin.exe into the archive and it should just work.

(Also, a new version of AppleWin is due out any day now ...)

Cheers,
Nick.

Michael AppleWin Debugger Dev

unread,
Sep 18, 2016, 1:54:37 PM9/18/16
to
On Sunday, September 11, 2016 at 7:04:44 PM UTC-7, Richard Thiebaud wrote:

> Virtual ][ (www.virtualii.com/) looks good but it is not free. A full
> license is $40.

"Looks" is subjective. Virutal II doesn't support the proper half-pixel shift, and the palette is absolutely horrible, but other then that, yes it is it pretty functional aside from the annoying habit of converting mounted .DSK into its proprietary file format.



D Finnigan

unread,
Sep 18, 2016, 2:21:16 PM9/18/16
to
Michael AppleWin Debugger Dev wrote:
>
> is it pretty functional aside from the annoying habit of converting
> mounted
> .DSK into its proprietary file format.
>

It rarely does this unless an emulated program clobbers the diskette.

I don't often use emulators, but when I do, this is the best one for 8-bit
Apples on Mac OS X.

And I've paid the full fee to get the full license.

--
]DF$
The Marina IP stack for Apple II--
http://marina.a2hq.com/

Antoine Vignau

unread,
Oct 7, 2016, 2:04:50 PM10/7/16
to
OpenEmulator and Virtual II perfectly emulate the Apple II hardware!

I am more in the IIgs side but these two can be trusted.

Antoine

Zellyn

unread,
Oct 11, 2016, 9:46:42 PM10/11/16
to
On Friday, October 7, 2016 at 2:04:50 PM UTC-4, Antoine Vignau wrote:
> OpenEmulator and Virtual II perfectly emulate the Apple II hardware!

Apparently not, since I'm seeing different output from each of them: https://imgur.com/a/XfupM
In Virtual ][, those dots are present, and flicker (but not if I emulate a II plus, interestingly).

The Mame one is just all weird.

If you want to try at home, the disk image is floatbus.dsk here: https://github.com/zellyn/a2audit/tree/master/floatbus

(I converted the "Have an apple split" first program to assembly.)

Zellyn

Zellyn

unread,
Oct 11, 2016, 9:48:13 PM10/11/16
to
On Sunday, September 11, 2016 at 9:13:00 PM UTC-4, Zellyn wrote:
> However, it appears that it puts the keyboard into some raw SDL mode, and I can't figure out how to bring up the MAME/MESS UI. I believe on Windows, it would be "scrl lock" or "ins", but neither of those keys exist on my MacBook keyboard.

For reference, and future searchability, I accidentally discovered that fn-delete will toggle putting mame into the mode where the keyboard controls can be brought up using "TAB".

sicklittlemonkey

unread,
Oct 12, 2016, 8:02:30 PM10/12/16
to
On Wednesday, 12 October 2016 14:46:42 UTC+13, Zellyn wrote:
> Apparently not, since I'm seeing different output from each of them: https://imgur.com/a/XfupM
> In Virtual ][, those dots are present, and flicker (but not if I emulate a II plus, interestingly).
>
> The Mame one is just all weird.

Yeah, MAME looks bad. I submitted my original floating bus routine to MAME
around 14 years ago, so not sure what it's doing now. The code I submitted
(which was also used in AppleWin) just fetches the appropriate video data
byte given a cycle count.

That was just to get some games working that wait for specific video data
by polling softswitches. The next step in properly rendering an image is to
sync the emulated video output with the emulated machine cycles. MAME didn't
do this, but it looks like Virtual II does - and now so does AppleWin.

As I mentioned elsewhere (as documented by Sather) video scanning is
slightly different in the II and //e.

Cheers,
Nick.

sicklittlemonkey

unread,
Oct 12, 2016, 8:03:33 PM10/12/16
to
On Wednesday, 12 October 2016 14:46:42 UTC+13, Zellyn wrote:
> On Friday, October 7, 2016 at 2:04:50 PM UTC-4, Antoine Vignau wrote:
> > OpenEmulator and Virtual II perfectly emulate the Apple II hardware!
>
> Apparently not, since I'm seeing different output from each of them: https://imgur.com/a/XfupM

Oh, could you post a pic of what OE looks like for that?

Cheers,
Nick.

Zellyn

unread,
Oct 12, 2016, 9:15:17 PM10/12/16
to
https://imgur.com/a/k8aSE - This is with a color monitor: it appears it leaves the color signal off for the entire screen render.

Michael J. Mahon

unread,
Oct 12, 2016, 9:42:11 PM10/12/16
to
Remember that the color burst, which is the only indication that a
composite signal contains color information, was never intended to switch
on and off rapidly.

Analog monitors typically take several frames to switch. Digital monitors
could do almost anything for a similar number of milliseconds.
--
-michael - NadaNet 3.1 and AppleCrate II: http://michaeljmahon.com

sicklittlemonkey

unread,
Oct 13, 2016, 12:01:59 AM10/13/16
to
On Thursday, 13 October 2016 14:42:11 UTC+13, Michael J. Mahon wrote:
> >> Oh, could you post a pic of what OE looks like for that?
> >
> > https://imgur.com/a/k8aSE - This is with a color monitor: it appears it
> > leaves the color signal off for the entire screen render.
>
> Remember that the color burst, which is the only indication that a
> composite signal contains color information, was never intended to switch
> on and off rapidly.
>
> Analog monitors typically take several frames to switch. Digital monitors
> could do almost anything for a similar number of milliseconds.

Heh, you've said this before and I was about to quote you. ; - )

So yeah, OE is showing the correct behaviour for an analog monitor.

AppleWin shows the same but in colour.

Cheers,
Nick.

Michael J. Mahon

unread,
Oct 13, 2016, 12:48:19 PM10/13/16
to
If a program "resides" in text mode, then momentarily switches into a
graphics mode for a fraction of a frame (small fraction?), an analog
monitor would remain in monochrome mode.

If, on the other hand, a program is in graphics mode almost all the time
and momentarily switches into text mode for fraction(s) of a frame, then an
analog monitor would remain in color mode.

The behavior of an (anachronistic) digital monitor is undefined. Some may
behave like analog monitors, but some may blank the image until a
"consistent" signal is detected. Others might try to track the
"instantaneous" color mode a line at a time (though I've never seen that
behavior).

sicklittlemonkey

unread,
Oct 13, 2016, 4:57:43 PM10/13/16
to
On Thursday, 13 October 2016 17:01:59 UTC+13, sicklittlemonkey wrote:
> So yeah, OE is showing the correct behaviour for an analog monitor.
>
> AppleWin shows the same but in colour.

I should clarify that I meant for Apple II scanning. For Apple //e scanning AppleWin actually shows something in between MAME and OE.

Have you tried MAME in Apple II emulation? It should be different.

Cheers,
Nick.

Peter Ferrie

unread,
Oct 14, 2016, 12:10:28 PM10/14/16
to
> I should clarify that I meant for Apple II scanning. For Apple //e scanning AppleWin actually shows something in between MAME and OE.
>
> Have you tried MAME in Apple II emulation? It should be different.

MAME has some progressive horizontal scanning, but it's a WIP.
We still can't show Crazy Cycles properly.

Charlie

unread,
Oct 14, 2016, 2:21:28 PM10/14/16
to
For what it's worth, I believe that if you are trying to make an
emulator as true to the Apple II as possible then it should emulate what
the Apple video output signal does and not what some "less than optimum"
monitor was displaying back in the day.
In other words graphics should be in color and text black and white with
no color fringe, even when the two are mixed.

I say "less than optimum" because I had a composite monitor in 1980 that
did a reasonably good job of displaying color in mixed mode with non
color fringed text below.

Charlie

Zellyn

unread,
Oct 14, 2016, 3:21:02 PM10/14/16
to
On Friday, October 14, 2016 at 2:21:28 PM UTC-4, Charlie wrote:
> For what it's worth, I believe that if you are trying to make an
> emulator as true to the Apple II as possible then it should emulate what
> the Apple video output signal does and not what some "less than optimum"
> monitor was displaying back in the day.

Well, OpenEmulator in particular also tries to emulate several specific monitors, so perhaps it should behave accurately on each of them :-)

Zellyn

Michael J. Mahon

unread,
Oct 14, 2016, 5:55:29 PM10/14/16
to
Do you remember what kind of monitor it was? That's extremely unusual
behavior.

And the Apple II video *leaves the color burst on* during the 4-line text
display under graphics modes, since NTSC is not speced for burst changes
within a frame.

sicklittlemonkey

unread,
Oct 15, 2016, 1:07:22 AM10/15/16
to
On Saturday, 15 October 2016 10:55:29 UTC+13, Michael J. Mahon wrote:
> Do you remember what kind of monitor it was? That's extremely unusual
> behavior.
>
> And the Apple II video *leaves the color burst on* during the 4-line text
> display under graphics modes, since NTSC is not speced for burst changes
> within a frame.

It's less the kind of video and more the kind of II.

My bet is that Charlie has a PAL IIe (like me) as it outputs white text in mixed modes.

Cheers,
Nick.

Charlie

unread,
Oct 15, 2016, 12:05:45 PM10/15/16
to
Yes it was a Samsung TV (13" I think) that was modified by a local
company to have a composite input. The modification was meant
specifically for Apple IIs.

> And the Apple II video *leaves the color burst on* during the 4-line text
> display under graphics modes, since NTSC is not speced for burst changes
> within a frame.

Not being a hardware guy I certainly won't argue that. All I know is
that it worked.
And yes I've been told by others that it couldn't have worked that way.
Usually sales people when I tried to buy another monitor.
Unfortunately, my monitor lasted only a year or so before something in
the high voltage section died.

Incidentally, the division between the graphics and text was not always
perfect. *Sometimes* you could see the top one or two scan lines of the
text portion flicker between color and black and white.

Charlie

Charlie

unread,
Oct 15, 2016, 12:08:02 PM10/15/16
to
It was an Apple II+ (not PAL) that I bought new in 1980 with the monitor.

Charlie


Brian Patrie

unread,
Oct 15, 2016, 6:11:38 PM10/15/16
to
On 2016-10-14 16:55, Michael J. Mahon wrote:
> And the Apple II video *leaves the color burst on* during the 4-line text
> display under graphics modes, since NTSC is not speced for burst changes
> within a frame.

Not the the II ever truly adhered to the NTSC spec. ;)

Michael J. Mahon

unread,
Oct 15, 2016, 6:55:41 PM10/15/16
to
True, as regards the details, such as no interlace. But the actual
"standard" to which Woz designed was compatibility with late-1970s analog
color TVs and monitors.

Monitor designs were slow (think a hundred milliseconds or more) to switch
from color to monochrome because it was not unusual for noise or
interference to interrupt the color burst for several lines or even a
frame. When switching from monochrome to color, it often required a frame
or two for the local chroma oscillator to phase lock to the color burst.

The time constant for switching modes was chosen to optimize for these
conditions.

A digital monitor could switch more quickly in the absence of noise, but
they didn't exist when the Apple II was designed.
0 new messages