19 replies to this topic

[Larry]

Does anyone have any experience running TBXL with the 65816 instead of the normal 6502?

With the 6502, TBXL works fine under Dos 2, Dos 2.5, and all versions of MyDos that I have tested (certainly 3.x and 4.5x).
With the 65816, TBXL continues to work under Dos 2.0/2.5, but fails to one degree or another under all versions of MyDos.

I realize that 65816's in our A8's are pretty scarce, but thought I'd ask (nothing ventured...).

-Larry

[drac030]

TBXL worked fine on my 65C816 Atari, when I had it. I of course only used it under SpartaDOS X, but that should not influence its general operation. Only BLOAD/BRUN instructions may fail under anything else than DOS 2.5, but this should always happen regardless of the CPU used.

TBXL is only sensitive to the math pack - if you replace the original math pack (in ROM) with the Fastchip, TBXL will fail.

[Larry]

Hi Drac030-

Thanks! The math pack might be a good clue. I'll check into that. The mode of failure is that it appears to load normally, setting up it's "splash screen" and continuing to load. But after loading, it returns to the Dup.Sys menu. But I'll also try SDX -- I haven't done that yet.

-Larry

drac030, on Fri Nov 25, 2011 9:58 AM, said:

TBXL worked fine on my 65C816 Atari, when I had it. I of course only used it under SpartaDOS X, but that should not influence its general operation. Only BLOAD/BRUN instructions may fail under anything else than DOS 2.5, but this should always happen regardless of the CPU used.

TBXL is only sensitive to the math pack - if you replace the original math pack (in ROM) with the Fastchip, TBXL will fail.

[Rybags]

Maybe MyDos does weird stuff with Basic ROM, like restoring it's state before running a program.

[Larry]

Hi Gary-

I'm not quite sure what you mean. It fails while loading TBXL -- returns to MyDos Dup.Sys. I probably did not make that clear.

-Larry

[Rybags]

OK... my theory was that MyDos might preserve PORTB state when doing Init/Run but probably not because it'd break a lot of stuff.

I thought it loaded totally, a partial load might mean that TBasic is unhappy about something and terminating the load process.

Are you doing this under emulation? If so, then you could upload an ATR and someone could take a look. Possibly TBasic requires certain memory requirements that MyDos isn't providing.

[David_P]

I'll have to dig out my 800xl with the FTe board, but I don't recall any issues with TBXL and a 65816.

[flashjazzcat]

TBXL works with MyDOS just fine, in my experience. Would a 65816 "silently" mess with PORTB in any way?

[Larry]

Yes, I've run TBXL with several different versions of MyDos and they all were just fine with 6502's.

There may be other issues.

@David_P -- I've since found that a "Sweet 16" board runs TBXL just fine, but this (different) board doesn't. So it's evidently not just the 65816. Today I'm going to trace the loads with Dos 2.0 (works) and MyDos (not).

@Drac030 -- this board uses the stock math pack.

-Larry

flashjazzcat, on Sat Nov 26, 2011 2:58 AM, said:

TBXL works with MyDOS just fine, in my experience. Would a 65816 "silently" mess with PORTB in any way?

[Larry]

Some progress -- MyDos 3.07 will load TBXL just fine. The 3.x versions could handle "large drives" but cannot navigate subdirectories.

Version 4.20 gives a clue. Midway through the load, it chokes and shows error 181 -- "Invalid address range for loading a binary file, end<beginning." After the error, it returns to Dup.Sys.

That address range certainly does not seem to be the case, since all of the load vectors are "legal."

-Larry

[bob1200xl]

Is this the system with 16MB of memory?

Bob

Larry, on Sat Nov 26, 2011 4:51 PM, said:

Some progress -- MyDos 3.07 will load TBXL just fine. The 3.x versions could handle "large drives" but cannot navigate subdirectories.

Version 4.20 gives a clue. Midway through the load, it chokes and shows error 181 -- "Invalid address range for loading a binary file, end<beginning." After the error, it returns to Dup.Sys.

That address range certainly does not seem to be the case, since all of the load vectors are "legal."

-Larry

[David_P]

I see. Sorry I can't be of more help then.

Obviously, the solution is to get more 65816 systems out in the wild...

[Larry]

Hi Bob-

No, it's not that one...

bob1200xl, on Sat Nov 26, 2011 7:34 PM, said:

Is this the system with 16MB of memory?

Bob

[Larry]

More success!

I remembered that I had an old (1980's) 6-sector ".EXE game loader" that works with most any Dos.Sys, and presents the typical 0-9 selection menu for selection. Works very well with MyDos 4.0 and 4.5 (so far). It does not put up the "splash screen" while loading, but appears to load everything correctly and goes straight to the blue screen "TURBO" or "READY". In testing several BASIC progs, TBXL works just fine, and will exit properly to Dos.

This is probably as close as I'm going to get to a "fix."

-Larry

[Kylev]

Larry,

I realize you have a valid work around, but did you try the obvious next step of creating a disk with just Dos.sys and naming TBXL as AUTORUN.SYS? I don't think it will save you anything. Just an additional step for completeness sake. Of course, you would lose the ability to exit to DOS.

No need to try, if you are not so inlcined.

[Larry]

Hi Kyley-

Thanks, I do appreciate the suggestion! I guess I hadn't mentioned it, but I had previously tried using TBXL as the Autorun.Sys for every version of MyDos that I had tested. It did work with version 3.x of MyDos, but not version 4.x, and I was really hoping for 4.x so that I could access subdirectories on my hard drive. You are quite right -- that is frequently a very good way to get around a program that can't be loaded from Dup.Sys.

As it stands now, I swap the hard drive to be D2: and boot from a floppy image with TBXL. MyDos 3.x probably works because it is much smaller than 4.x. IIRC, it is even smaller than Dos 2.0.

-Larry

Kylev, on Sun Nov 27, 2011 8:01 PM, said:

Larry,

I realize you have a valid work around, but did you try the obvious next step of creating a disk with just Dos.sys and naming TBXL as AUTORUN.SYS? I don't think it will save you anything. Just an additional step for completeness sake. Of course, you would lose the ability to exit to DOS.

No need to try, if you are not so inlcined.

[Larry]

Yes, it's too bad that there haven't been more "Sweet-16" or "Turbo 816" type boards produced. I don't think that FTE ever produced/sold very many of them. I remember hearing that there were "issues" (?) with some of those boards, but I don't know what those issues were. OTOH, there isn't much "bang for the buck" by using a 65816 at the same clock speed except to find out how many things "break" when the 6502 is swapped out.

-Larry

David_P, on Sat Nov 26, 2011 9:02 PM, said:

I see. Sorry I can't be of more help then.

Obviously, the solution is to get more 65816 systems out in the wild...

[rost]

you can try to rename the autorun.sys to AUTORUN.AR0 for mydos...

[Larry]

Thanks, I hadn't tried that, but it fails, too -- goes back to Dup.Sys.
-Larry

rost, on Mon Nov 28, 2011 8:09 AM, said:

you can try to rename the autorun.sys to AUTORUN.AR0 for mydos...

[bob1200xl]

The 65816 does not generate PH02 timings for you, where the 6502 does. So, the interaction between the other chips and the 65816 are just enough different to cause problems with some ANTICS on the Sweet16. I think Larry's problem is similar, only with POKEY timings.

Bob

Larry, on Sun Nov 27, 2011 11:01 PM, said:

Yes, it's too bad that there haven't been more "Sweet-16" or "Turbo 816" type boards produced. I don't think that FTE ever produced/sold very many of them. I remember hearing that there were "issues" (?) with some of those boards, but I don't know what those issues were. OTOH, there isn't much "bang for the buck" by using a 65816 at the same clock speed except to find out how many things "break" when the 6502 is swapped out.

-Larry

David_P, on Sat Nov 26, 2011 9:02 PM, said:

I see. Sorry I can't be of more help then.

Obviously, the solution is to get more 65816 systems out in the wild...