MiSTer FPGA Forum

That's what always happens when I curl into a tearful ball of desperation on the floor in front of my computer: it gets unstuck. I just generated my first FAT12 file system from scratch with a file on it, and not even Norton Utilities finds it suspicious anymore. To be continued!

Anyone here up to speed on how to exactly compose a FAT12 allocation table? I'm sooo very close right now to getting PC DOS 2.00 booting correctly, but my implementation is ever so slightly off..

I'll dig into the code for a bit.. If there really are things in the core that are unreachable from SYSCTL, I'll submit a PR to get that rectified. The C program looks simple enough. I'll just have to hunt for the twiddly bits in the core's code fore a bit because I don't know any HDL.

Something I noticed is that these settings only flip once from within DOS and I need to reset the the core for them to work again. Not something I spent a lot of time on though so I may be wrong and I usually reset my core anyway because of the single game VHD's I use. I haven't had any real issues developing DOSContainer.

Those presets are just combinations of clock and cache settings. Those are all available from SYSCTL.

Both the length and the contents of the sequence are different every time. I guess it's a bit late to report a bug to IBM now, but this does seem to be one. This may become a thing when the disk gets filled to near capacity.

Norton Utilities doesn't like these disks much either. Ah well, the things we unearth from the depths of digital archaeology.

Flynnsbits's answer from a year or 2 ago is still as valid as it was when he posted it. VESA won't work on this core in its current state.

More historical computers would still be very nice.. like the actual PDP11, early IBM mainframe, VAX, heck.. Burroughs or Data General big iron would fit easily and be interesting to run ALGOL60 on. But I guess that's just me being interested in computing and conservation rather than only playing video games.

For anyone interested, I gathered A LOT of DOS stuff here:

ftp://ftp.area536.com/platforms/PC

Have you considered lending to a serious museum?

Those don't affect the sound. You can flip the device to emulate a General MIDI synth using SoundFonts. That's departing from the MT32 though. The original MT32 was what it was and sounded the same whenever you got it to work.

The Pi acts like an MT-32. If the game supports and detects it properly, there's nothing else to set. Some games expect a certain MT32 ROM version but that depends on the game.

Gearing up nicely. Builds are now generated automatically for platforms other than just my local laptop. Unfortunately I'm not (yet) able to do Mac builds from Linux due to licensing issues from Apple.

More info here: https://www.area536.com/posts/doscontainer-rust-cicd/.

That'll be "collateral" from the analysis part of the tool. Until then you could use GNU Mtools to do this.

Not that this will work with MiSTer, but some may find this interesting:

https://www.furygpu.com/

It's a complete GPU implementation that would equate a mid-90's graphics card.

Not to hijack this one, but the DOSContainer Bash version with the YAML's for bare MS-DOS could do what you're looking for. Edit the YAML, run script, and you have a clean DOS with some toggles like drivers and memory mgt and the like. Shoot me a PM if you are interested.

Make a throwaway copy of your VHD and strip down the CONFIG.SYS. Start up from it and run MEMMAKER and follow the steps. That'll change your CONFIG.SYS for you, hence the throwaway VHD. Note the line that loads EMM386.EXE. Copy that down and into the VHD that has your menu system. That'll probably fix it in a reasonably optimized way.

..right.. PCEm does similar but subtly different things yet again. It's the FORMAT utility that puts the weird bytes there. No idea why though..

Ah.. fudge.. Just noticed. I tried the PCXT core to run PC-DOS but that wouldn't boot of the 360k floppy image. So I tried AO486, and that created a disk with the wrong geometry using 63 sectors per track. Those didn't exist yet when DOS 2.00 was around. Back to the drawing board (and PCEm) for some more testing.

I've never written a line of Verilog in my life, but compilation of cores really isn't that hard. It's just pushing a button and waiting.

Emubastard wrote: ↑Sun Mar 24, 2024 5:52 pm

VHD's are generated manually by Windows 11 Disk Management utility so they are indeed real VHD's. Easier to work on when I have issues to fix, which there are many lol.

-Emubastard

Wouldn't be the first time MS don't follow their very own specs. Thanks for letting me know.

DOSContainer Rust just wrote its first file (IBMBIO.COM) to a VHD that was bootstrapped completely from scratch, zero external dependencies. Things still aren't ready for prime time and the code surrounding the writes is messy, but I'm very slowly getting there.

DR-DOS is better at freeing conventional memory than MS ever was. That may help.