Proper OS65D V3.2/3.3 disk image anywhere?

Post Reply
tanteju
Posts: 17
Joined: Mon Dec 24, 2018 12:28 am

Proper OS65D V3.2/3.3 disk image anywhere?

Post by tanteju »

Hi,

Before anyone replies: "Look on the webpage", let me explain the background:

In my effort to get my EC-65/Octopus 65/Samson 65 (all name the same computer) running again with OS65D I'm trying to find out, where the original images had to be patched, for this getting working.

And regardless what I try, I always run into a BRK somewhere. Unfortunately not really 100% replicable, so there are tons of variables. To exclude bad memory I've already moved all RAM to a modern 128 kByte SRAM chip with the same results.

So, going through all the documentation I have and that is on the web page I thought going back to 3.2, because there is a commented disassemble available, might be an option. But all the images I found have in the boot page (2200) in the beginning a JSR to a routine in 2Exx. As far as I know, this should not be the case for 3.2, as that routine was introduced with 3.3 or maybe with 3.3+overlays.

Therefore I'm suspicious, that the image was somehow back-engineered from a newer version. And if there is something down in the weeds of 3.3 that is producing my issues I fear, it might in that type of image as well. Hence my question if someone has a really original version of 3.2 available?
MK14HAK
Posts: 356
Joined: Wed Mar 16, 2011 1:49 am
Location: New Zealand

Re: Proper OS65D V3.2/3.3 disk image anywhere?

Post by MK14HAK »

Welcome tanteju. The images on OSIWeb are "proper". There are two major versions of 65D32. The Tutorial 2 disk of the 65D33 set is actually v3.2 and the 3.2 NMHz version. This loads code at $2E79 to determine the drive step rate and either sets $267B to $31 or $62. Some disks also load a true ASCII kybd driver for v3.2 also from TK6. @$3180 or $3200 from memory.You could bypass the call to 2E79 and just set $267B if you like.
From your other posts, I understand you have the kernel working for R/W to a real floppy. Are you booting to the Kernel and then trying to launch Disk BASIC.
What kernel commands or routines are crashing ?
Are you able to get the Assembler or Exmon to run ?
Mike
600RevB:16K,2MHz,64x32,470,CEGMON
SuperKit:502,540B,542B,CEGMON, 8" and 5" FDDs
Cards:PE IO,6522 D-A-D, AY3-8910,ProgramGraphics,Color,UK101
WIP:HexDOS,FDD Emulator
tanteju
Posts: 17
Joined: Mon Dec 24, 2018 12:28 am

Re: Proper OS65D V3.2/3.3 disk image anywhere?

Post by tanteju »

MK14HAK wrote: Thu Jan 17, 2019 9:32 pm Welcome tanteju. The images on OSIWeb are "proper". There are two major versions of 65D32. The Tutorial 2 disk of the 65D33 set is actually v3.2 and the 3.2 NMHz version. This loads code at $2E79 to determine the drive step rate and either sets $267B to $31 or $62. Some disks also load a true ASCII kybd driver for v3.2 also from TK6. @$3180 or $3200 from memory.You could bypass the call to 2E79 and just set $267B if you like.
From your other posts, I understand you have the kernel working for R/W to a real floppy. Are you booting to the Kernel and then trying to launch Disk BASIC.
What kernel commands or routines are crashing ?
Are you able to get the Assembler or Exmon to run ?
Hi MK14HAK,
For the NMHZ Version I see the need for the added subroutine, but the plain 3.2 should not have that, according to http://osiweb.org/manuals/OS65D_Dissasembly_Manual.pdf
My 3.0 disassembly (posted that in another thread) does not have that as well.

I have the Kernel running. I can read and write sectors (writing on the Gotek does not really work, that's a different topic). Starting BASIC or Assembler sometimes works, but running BEXEC* never works.

I now have produced a 3.3 disk 5 image, that at least gives me some reproducable errors.
Booting Kernel and Basic will show Error #7 when it comes to BEXEC*. When I exit Basic and type "LOAD BEXEC*" in the kernel, BEXEC* seems to be loaded into RAM. But every time a write operation in video RAM (E800-EFFF) occurs, because suddenly on the right side of the screen a "(" (open parentheses) occurs. When I try to go back into Basic and run that program either I get an error in Line 100 or it crashes and I'm in monitor program. Listing never works (yes, I've enabled the command before).

As I have one diskette, where fragments are still readable, and tried with that disk loading BEXEC* I have exactly the same issue with the character appearing.

As nothing in the Kernel should write directly to Video RAM (all Video handling should be done via EPROM subroutines), I wonder if I missed something to patch in the Kernel, maybe some indirect write which is not that easy to find? Hence my idea, that this is linked to the modifications not included in the "normal" 3.2 disasm.

Assembler starts, but when I have a file loaded and I try to show it, the system just hangs.

I think I need to find a way to get data directly to my Mac or so, so that I could copy the data from the working tracks of the one floppy that still partly works. But track 0 is for sure gone there.
MK14HAK
Posts: 356
Joined: Wed Mar 16, 2011 1:49 am
Location: New Zealand

Re: Proper OS65D V3.2/3.3 disk image anywhere?

Post by MK14HAK »

The Software Consultants disassembly is of an 8" DOS32 that I have not found an equivalent image for. Don't regard it as the "proper" 3.2 DOS as such. Although it mentions NMHz it does not make any call to the $2E79 NMHz routines and they have not mentioned these. DOS32 NMHz can be found on OSIWeb and Marks website and is a proper OSI release of OS65D32.2 DOS.Also the Computer Consultants disassembly has minor errors and will not assemble and run as per the listing. I have been there and tried. I does not work. But as far as being useful for your patches to video, keyboard and disk routines it will be fine and is a great reference.
There are direct writes to OSI areas that will not be compatible with your EC-65/Octopus 65/Samson 65. eg BFXX, DE00, F4XX,FBXX,CFXX. You have probably found these.
I would be getting a 3.2 Kernel sorted 100% first with all your Monitor calls.Ensure all commands work. Then maybe create a small MC routine using keyboard and video routines eg STROUT call. (Assembler can be patched easily) Ensure you have no "((" appearing any where with screen scrolls etc and it doesnt drop to DOS with Error #7. Then on to 3.2 BASIC. Avoid Bexec* for starters.Set 2321,2=02. Unlock from the command line and write a " hello world" program and test commands...

It would be very handy if you could run WinOSI in a virtual XP box on your Apple. Have you got a serial interface between the two.
Last edited by MK14HAK on Fri Jan 25, 2019 7:29 am, edited 1 time in total.
600RevB:16K,2MHz,64x32,470,CEGMON
SuperKit:502,540B,542B,CEGMON, 8" and 5" FDDs
Cards:PE IO,6522 D-A-D, AY3-8910,ProgramGraphics,Color,UK101
WIP:HexDOS,FDD Emulator
tanteju
Posts: 17
Joined: Mon Dec 24, 2018 12:28 am

Re: Proper OS65D V3.2/3.3 disk image anywhere?

Post by tanteju »

MK14HAK wrote: Fri Jan 18, 2019 1:18 am The Computer Consultants disassembly is of an 8" DOS32 that I have not found an equivalent image for. Don't regard it as the "proper" 3.2 DOS as such. Although it mentions NMHz it does not make any call to the $2E79 NMHz routines and they have not mentioned these. DOS32 NMHz can be found on OSIWeb and Marks website and is a proper OSI release of OS65D32.2 DOS.Also the Computer Consultants disassembly has minor errors and will not assemble and run as per the listing. I have been there and tried. I does not work. But as far as being useful for your patches to video, keyboard and disk routines it will be fine and is a great reference.
There are direct writes to OSI areas that will not be compatible with your EC-65/Octopus 65/Samson 65. eg BFXX, DE00, F4XX,FBXX,CFXX. You have probably found these.
I would be getting a 3.2 Kernel sorted 100% first with all your Monitor calls.Ensure all commands work. Then maybe create a small MC routine using keyboard and video routines eg STROUT call. (Assembler can be patched easily) Ensure you have no "((" appearing any where with screen scrolls etc and it doesnt drop to DOS with Error #7. Then on to 3.2 BASIC. Avoid Bexec* for starters.Set 2321,2=02. Unlock from the command line and write a " hello world" program and test commands...

It would be very handy if you could run WinOSI in a virtual XP box on your Apple. Have you got a serial interface between the two.
I HAD IT BOOTING. COMPLETELY INTO THE BEXEC* MENU.

It seems, the issue is not where I searched. It is not the software. Something in the hardware is broken. Either a bus driver or a capacitor.
Post Reply