Apple IIGS MiSTer Core Progress

[ExCyber]

I was just messing around trying to get a feel for it, so I wasn't keeping close track of everything, but two games that failed are Ancient Land of Ys (PRODOS fails from HDD with "No SYSTEM/P16 file found $000E", but works if run from floppy; YSBOOT.SYS16 asks for the data disk and crashes if it's inserted) and Arkanoid ("CAN'T LOAD TAITO.ARCADE 0000" from HDD, launches from Finder if whole-disk copied to RAM5 but with severe graphical glitches, works normally if booted from RAM5). Out of This World initially complained about TOOL025 missing, then just relaunched Finder, but eventually started working after I messed with some other stuff and I'm not sure why (maybe something needed a reboot?).

Dragging a disk to a GS/OS folder doesn't work for me ("Disk icons can be dragged only onto other disk icons."). Might this have been enabled by some third-party software?

[cathrynmataga]

I have real hardware for IIGS, and I find running anything Prodos16, which is most games like Ys, pretty hit or miss from GS.OS. You can usually spot real GS.OS apps because the desk accessories from the GS.OS show up in the app. Some things that do install to GS.OS that I've been tinkering with. Hypercard, Beagle Write and Beagle Draw, Orca C, Jam Session. Some of these have installs, Jam Session wants one of its directories (I think the one with its sound banks) in the root of the hard drive. ProDos 8 stuff seems a bit more reliable than GS.OS from hard drive. Software installed to GS.OS seems to load faster than Prodos16 which can have some silly slow loading times.

[thorr]

Thanks for your reply ExCyber! To keep it short, I am only quoting you and not my original post...

That's a cool hack, but I think it depends on using the actual SmartPort firmware and protocol (and would also be limited to a floppy drive data rate?). The AppleWin HDD is a very simple device that just supports the ProDOS firmware calls.

It's not a hack. It is taking advantage of the SmartPort functionality that is built into the external floppy connector. The speed is significantly faster than floppy speeds, but wouldn't be as fast as AppleWin HDD probably. The FloppyEmu emulates real hardware, both floppy drives and SmartPort HDD's which natively support multiple drives.

There isn't really one agreed-upon "max", so RAM size might need to be an option for compatibility reasons. 4 MB was the official limit for RAM expansion size, while 8 MB became possible with later RAM chips (i.e. the vast majority of IIgs software probably never encountered this much RAM during its support lifetime). More than that seems to require patching GS/OS and decoding banks 80-DF as RAM (as far as I know, not readily possible without an accelerator on an actual IIgs).

When I said to max it out, I meant whatever the standard compatible amount would be which would be 4 MB in this case. Unless there is some compelling reason to go beyond that, I would stick with 4 MB.

The serial port is a Zilog SCC (Z8530), which I think needs to be written. The only existing core I know of that would need this chip is MacPlus, and it doesn't seem to be there.

Understood, so hopefully it can be written eventually, and can be added to both cores. How cool would it be to use an ImageWriter II printer with a MiSTer IIgs core?!

Do you mean something like the "Composite Blend" option in the MegaDrive core? The composite output on IIgs is just RGB put through an off-the-shelf encoder chip.

I am not sure, maybe. I mean emulate the look of composite video for legacy games that take advantage of the way the colors interact with each other (artifacts) to produce more color. See this page for exactly what composite video causes: https://scruss.com/blog/2018/05/15/appl ... -graphics/ and then the behavior that a composite monitor does to make it look good instead of like it does on an LCD. This video is along the lines of what I am talking about:https://www.youtube.com/watch?v=niKblgZupOc I am pretty sure the AppleWin emulator already has the functionality built in for the IIe (NTSC vs RGB). I don't know though. Maybe me just using my NTSC TV will automatically provide the look I am after even when using component. I don't think so though.

The issue with stuff like this is that it makes the core bigger for very little benefit. Maybe a Mockingboard is tiny, though (it's been a while since I've looked at the AY, but I want to say it's basically three counters in a trenchcoat)? This would probably also wind up needing an implementation of the "Slotmaker" chip, which controls the mapping of internal vs. expansion slot ("Your Card") hardware.

The amount of benefit is great, in my opinion. I guess it depends on what each individual wants, and I know that I would want it 100%. If I owned a IIe back in the day with a Mockingboard, and later upgraded to a IIgs, for sure that Mockingboard would be moved to it. I might even buy one for the games that support it anyway if I only ever owned a IIgs. The Mockingboard is already in the IIe core, so it just needs to be duplicated to make two of them (two in the IIe core and two in the IIgs). I guess my vision of the IIgs core is eventually a complete core with everything it can do (Slotmaker, Smartport, etc.). Otherwise it would be an incomplete core and might have compatibility issues. I am a perfectionist when it comes to cores I really care about and will bring to light anything I think is missing, but I am just a guy on the internet hoping for the best outcome with no say on the matter other than caring about it. If there is no point to making something (like maybe Smartport functionality) because it doesn't add anything useful and doesn't break anything, then I am ok with it being missing.

I'm not super-familiar with IIgs software, but I think most games are designed to be booted. If they use a standard ProDOS loader (either originally or because they were cracked) they might not care too specifically about the disk format and perhaps could be mapped to a RAM disk or similar. I don't know what it would take to build something like WHDLoad for GS/OS, or if anyone's attempted/done that.

There has been further useful posts about this. I would just add that system files, etc. need to be in the right places for it to work. ProDOS is pretty flexible and it might depend on if the software is hardcoded to a certain path or if it can work with relative paths.

Thanks again!

[ExCyber]

Likewise, trimming quotes and just replying to the parts I'm prepared to at the moment.

It's not a hack. It is taking advantage of the SmartPort functionality that is built into the external floppy connector. The speed is significantly faster than floppy speeds, but wouldn't be as fast as AppleWin HDD probably. The FloppyEmu emulates real hardware, both floppy drives and SmartPort HDD's which natively support multiple drives.

I was mixing up two different descriptions of the HDD emulation; the "hack" part I was thinking of is supporting a 2 GB drive on the built-in firmware (by formatting it as an HFS volume rather than a ProDOS volume), but even that might actually be technically supported by the firmware API. An emulated SmartPort HDD is faster than floppy in the sense that seeking is practically instantaneous, but as far as I know the raw data transfer over SmartPort is limited to roughly the same speed as a SmartPort 3.5" drive (which in turn is roughly twice the transfer rate of a Disk II). This seems to fit with a YouTube video I saw of the Floppy Emu taking over a minute to boot GS/OS as a SmartPort HDD (although I don't know what extensions/mods were being loaded).

When I said to max it out, I meant whatever the standard compatible amount would be which would be 4 MB in this case. Unless there is some compelling reason to go beyond that, I would stick with 4 MB.

From what I've read, it seems like the main use of more RAM is to put OS and application files on a RAM disk for extra speed. The MAME driver documentation says that 1 MB of expansion RAM (for 2 MB total, since MAME defaults to a ROM3 machine) is enough to run "all known software", although I'd bet there are some IIgs power users out there who would disagree with that assessment. I initially thought that the RAM disk use case could be a big deal because I assumed that slot card DMA was limited to the legacy RAM/speed, but in fact they can DMA to fast RAM, which was actually part of the compatibility concern with larger RAM expansions. 4 MB of expansion RAM with a fast storage controller in slot 7 seems like it would be a pretty good setup. The current AppleWin HDD is not optimal for this (no DMA, but still halts the CPU to do the framework transfer).

I mean emulate the look of composite video for legacy games that take advantage of the way the colors interact with each other (artifacts) to produce more color. See this page for exactly what composite video causes: https://scruss.com/blog/2018/05/15/appl ... -graphics/ and then the behavior that a composite monitor does to make it look good instead of like it does on an LCD. This video is along the lines of what I am talking about:https://www.youtube.com/watch?v=niKblgZupOc I am pretty sure the AppleWin emulator already has the functionality built in for the IIe (NTSC vs RGB). I don't know though. Maybe me just using my NTSC TV will automatically provide the look I am after even when using component. I don't think so though.

That makes sense, but I think it's more of an enhancement to the framework video processing than something that belongs to a specific core. There's already a pair of filters that I guess are meant to blur somewhat like a composite CRT, but they look weird on my monitor and don't seem to generate artifact colors (or I've forgotten which details to look for).

There has been further useful posts about this. I would just add that system files, etc. need to be in the right places for it to work. ProDOS is pretty flexible and it might depend on if the software is hardcoded to a certain path or if it can work with relative paths.

Yeah, I probably need to RTFM, or maybe see if some magazine had a nice tutorial article back in the day, since modern guides are more emulator-centric. I only know the very basics of ProDOS, and most of that was with ProDOS-8, not ProDOS-16.

[cathrynmataga]

Out of This World initially complained about TOOL025 missing, then just relaunched Finder, but eventually started working after I messed with some other stuff and I'm not sure why (maybe something needed a reboot?).

Btw, if you see a game complaning about TOOL025 or TOOL-anything, that means you need to copy the TOOL025 from your SYSTEM:TOOLS directory on your game disk to :SYSTEM:TOOLS on your boot drive. This is usually a good sign, that the app is meant to work with GSOS and might work with a little fiddling. Not sure what happened with Out of this World, but maybe they have some code in there that copied the TOOL file.

[thorr]

An emulated SmartPort HDD is faster than floppy in the sense that seeking is practically instantaneous, but as far as I know the raw data transfer over SmartPort is limited to roughly the same speed as a SmartPort 3.5" drive (which in turn is roughly twice the transfer rate of a Disk II). This seems to fit with a YouTube video I saw of the Floppy Emu taking over a minute to boot GS/OS as a SmartPort HDD (although I don't know what extensions/mods were being loaded).

Agreed. I have an Apple IIc with a Floppy Emu, and I know for sure that after I transferred all my programming stuff to a SmartPort HDD, it loaded much faster than it did from the 140K floppy images. It is quite possible it is similar to a 3.5" floppy in performance only faster due to the seek times. It is a welcome and very noticeable performance improvement over 5.25" floppies.

From what I've read, it seems like the main use of more RAM is to put OS and application files on a RAM disk for extra speed. The MAME driver documentation says that 1 MB of expansion RAM (for 2 MB total, since MAME defaults to a ROM3 machine) is enough to run "all known software", although I'd bet there are some IIgs power users out there who would disagree with that assessment. I initially thought that the RAM disk use case could be a big deal because I assumed that slot card DMA was limited to the legacy RAM/speed, but in fact they can DMA to fast RAM, which was actually part of the compatibility concern with larger RAM expansions. 4 MB of expansion RAM with a fast storage controller in slot 7 seems like it would be a pretty good setup. The current AppleWin HDD is not optimal for this (no DMA, but still halts the CPU to do the framework transfer).

I can't speak about this too much because I have never used a IIgs. However, I did use RAM disks on the IIc and Laser128EX that I used to have and it was awesome performance. This could be useful for some programs where everything is loaded onto the RAM disk and then runs from there. Also, Appleworks could take advantage of the extra RAM. I am all for the fastest and most compatible HDD solution.

That makes sense, but I think it's more of an enhancement to the framework video processing than something that belongs to a specific core. There's already a pair of filters that I guess are meant to blur somewhat like a composite CRT, but they look weird on my monitor and don't seem to generate artifact colors (or I've forgotten which details to look for).

I am not sure I agree. The Apple II has some very specific color artifacting behavior. It's possible that the RGB mode automatically takes care of this, but I don't know. We can wait and see on this one. It's not a day one requirement by any means. And thinking about it, probably using a composite output out of the MiSTer with an adapter paired with my Apple IIc monitor will have the intended affect automatically. It will be interesting to compare the same programs with my Apple IIc on the same monitor.

Yeah, I probably need to RTFM, or maybe see if some magazine had a nice tutorial article back in the day, since modern guides are more emulator-centric. I only know the very basics of ProDOS, and most of that was with ProDOS-8, not ProDOS-16.

Again, I have no IIgs experience. My knowledge is only on the ProDOS-8 side of things. I am looking forward to eventually (hopefully) playing with and learning all about the IIgs!

[ExCyber]

I am not sure I agree. The Apple II has some very specific color artifacting behavior. It's possible that the RGB mode automatically takes care of this, but I don't know. We can wait and see on this one. It's not a day one requirement by any means. And thinking about it, probably using a composite output out of the MiSTer with an adapter paired with my Apple IIc monitor will have the intended affect automatically. It will be interesting to compare the same programs with my Apple IIc on the same monitor.

What I'm saying is that the IIgs is not "the Apple II" in that sense. The classic bitstream only exists internally between two of the custom chips, one of which digitally converts it to RGB. So while the composite output necessarily has artifacts, I don't know that there's anything special about them compared to any other RGB system with a composite output.

That being said, it's been a long time since I've seen a IIgs in person, and I don't remember anything in particular about the monitor. I do have a CRT TV that still worked when I pulled it off the shelf a year or two ago. Maybe I should just buy a IIgs and see what happens instead of speculating. But then I'd probably also feel compelled to throw a couple hundred bucks at BMOW to avoid needing to find working drives/disks and ADB keyboard/mouse, maybe buy a SCSI card and BlueSCSI, a modern power supply... it could get expensive pretty fast.

[Flandango]

I just got a IIgs restored with it's rgb monitor but I don't have a crt tv and the monitor doesn't handle composite video.
The RGB monitor does not produce the composite artifacts like a composite tv/monitor would.
The IIgs does have a composite output for low res graphics (aka legacy II software).
I don't know what would happen if you were to covert the RGB output signal (from the core that is) to composite monitor what it would do.

If there is anything I can help with the IIgs, let me know and I'll see what I can do....I have recently become WAY too familiar with it's signaling and what not trying to troubleshoot a non working motherboard, sound and sound ram.....

[thorr]

What I'm saying is that the IIgs is not "the Apple II" in that sense. The classic bitstream only exists internally between two of the custom chips, one of which digitally converts it to RGB. So while the composite output necessarily has artifacts, I don't know that there's anything special about them compared to any other RGB system with a composite output.

That being said, it's been a long time since I've seen a IIgs in person, and I don't remember anything in particular about the monitor. I do have a CRT TV that still worked when I pulled it off the shelf a year or two ago. Maybe I should just buy a IIgs and see what happens instead of speculating. But then I'd probably also feel compelled to throw a couple hundred bucks at BMOW to avoid needing to find working drives/disks and ADB keyboard/mouse, maybe buy a SCSI card and BlueSCSI, a modern power supply... it could get expensive pretty fast.

Thanks ExCyber and Flandango. I have another idea too. I have an Apple IIc with composite output and Apple IIc composite monitor, and I have a composite adapter for my MiSTer, and I have a TV with composite and component video (effectively equivalent to RGB), and we have the Apple IIe core. I can take pictures of the Apple IIc monitor and TV to do comparisons:

• Apple IIc connected to Apple IIc monitor via composite
• MiSTer IIe connected to Apple IIc monitor via composite
• Apple IIc connected to TV via composite
• MiSTer IIe connected to TV via composite
• MiSTer IIe connected to TV via component

This should tell us most of the current story hopefully.

And ExCyber, you are making me all the more tempted to get a IIgs because I already have a FloppyEmu, ADB keyboard/mouse, BMOW USB wombat that allows USB keyboards and mice to connect to ADB, BMOW Floppy Emu, and BlueSCSI, plus an Apple IIc monitor and TV that could be used for RGB with an adapter. I am still resisting that temptation because I would rather have a IIgs MiSTer core actually.

I just got a IIgs restored with it's rgb monitor but I don't have a crt tv and the monitor doesn't handle composite video.
The RGB monitor does not produce the composite artifacts like a composite tv/monitor would.
The IIgs does have a composite output for low res graphics (aka legacy II software).
I don't know what would happen if you were to covert the RGB output signal (from the core that is) to composite monitor what it would do.

If there is anything I can help with the IIgs, let me know and I'll see what I can do....I have recently become WAY too familiar with it's signaling and what not trying to troubleshoot a non working motherboard, sound and sound ram.....

Thank you for answering the question about the behavior of the RGB monitor not producing the composite artifacts. Hopefully my experiment will answer the question about converting the RGB output to a composite monitor because the MiSTer IIe to Composite is basically that experiment, I think.

[rcade]

Yes the RGB output of the GS definitely does simulate the artifact color! The Apple II portion would be black and white without that, and it is not. All games look normal.

It's actually the composite output that is not very useful, since the IIGS colors do not work right on that connection.

[Flandango]

Ah ok...good to know. I was under the impression that while it was in color, some of the artifacts that were related to the composite signal didn't show up on the rgb monitor so things looked...off.
I myself didn't have anything to compare since the monitor I have (A2M6014) doesn't have composite input and it will be pointless for me to try the composite out on any of my lcd screens.

[cathrynmataga]

I only have RGB hooked up now, but I believe on the monitor jack, it does color IF you set Control Panel/Display/Type = Color, AND you're not in an all text mode. I did see color on a monitor connected directly myself. In AppleSoft, if you type GR, you'll see text with fringes on the bottom four lines, but no color fringes in the regular text mode. Maybe is monitor dependent?

[thorr]

You can actually turn that color fringing on the text off and on in the Apple IIe core in the core menu. It's a special MiSTer core feature that a normal Apple IIe can't do. It is normal to have the weird colors on the text when in graphics modes, even on my Apple IIc with composite monitor. It will be interesting to see how it looks on screen compared with RGB/component.

It would be great to see some pictures of the Apple IIgs RGB monitor in action with Apple IIe programs. I am planning to do the comparison pictures I mentioned above hopefully by this weekend. I will take some photos of Oregon Trail where there is a picturesque scene and also some HGR and GR graphics in BASIC with some text at the bottom, and some King's Quest for double hires.

[Flandango]

Here is a picture I just took in Oregon Trail....is the detail ok? Let me know and I can grab a couple more and zip them up.
I took with my S24 phone so hopefully it wasn't "AI" altered.

[thorr]

Thanks! Yes, that definitely works. I forgot to mention I have a Mac LCIII with an Apple IIe card in it. The Mac LCIII effectively uses a VGA monitor, and I can tell you that the above picture looks way better than the Mac does. Your picture has that grain to it that disappears on the Mac. Also, I can see the color fringing in the M in March. I think I have some photos from my IIc already, so I will try to find and post one of those for comparison...

[thorr]

It's not exactly the same photo, but close enough. Here it is from my Apple IIc monitor...

gbQYdyO.jpeg (1.15 MiB) Viewed 2 times

[Flandango]

I had first noticed that fringing in the M on the title screen and couldn't tell if that was intentional or not. Was trying to take a picture of it zoomed in, but my phone kept "fixing" it and making a solid M.
As soon as I can, I'll grab the couple of other pictures I took (not many) and bring em over to my pc to zip them up instead of posting huge pictures.

[thorr]

Thanks! I have a ton more pictures I took previously for akeley's project so I can post more too. I think from even these two pictures that it is clear that the color artifacting is not as present in the IIgs version on the RGB monitor. For example the buildings on the right look quite different. Also, the blue is very different which is interesting.

[Flandango]

Yeah I noticed that too.
I would still like to get a hold of at least a tube tv to at least try out the composite output for comparison.
My next door neighbor use to have a small tv so will hit her up to see if she still has it buried somewhere.

[Flandango]

Here are a couple more pictures.
Please pardon the rbg light reflection you see on the left hand side. I didn't notice them on the small screen of the phone.

[thorr]

Thanks! I had comparable pictures to most of those. See the comparison here:
https://imgur.com/a/PbZFA8P

[thorr]

Here are some more examples on the Apple IIc composite monitor: https://imgur.com/a/Wy11M9P
You can click on these to make them bigger.

[ExCyber]

I only have RGB hooked up now, but I believe on the monitor jack, it does color IF you set Control Panel/Display/Type = Color, AND you're not in an all text mode. I did see color on a monitor connected directly myself. In AppleSoft, if you type GR, you'll see text with fringes on the bottom four lines, but no color fringes in the regular text mode. Maybe is monitor dependent?

You can actually turn that color fringing on the text off and on in the Apple IIe core in the core menu. It's a special MiSTer core feature that a normal Apple IIe can't do. It is normal to have the weird colors on the text when in graphics modes, even on my Apple IIc with composite monitor. It will be interesting to see how it looks on screen compared with RGB/component.

I stumbled upon an Apple tech note recently suggesting that they decided to accept the fringing on text in GR mode because most displays couldn't respond to the absence of the colorburst fast enough to actually switch to monochrome mode mid-frame. The "color killer" is in a different place in IIgs, but it seems like it would run into the same limitations when using a composite display.

Making this switchable isn't a big deal in a MiSTer core because the framework uses RGB and we control the color decoder, so we can make our own decision to bypass it.