TI99/4a OS versions

[+videofx]

6 minutes ago, senior_falcon said:

That's an interesting observation. I had forgotten that there were some slight differences in the roms.

@videofx

Try this simple program on your two consoles (any XB will work fine) and report the results. This is not super fast, so patience is necessary.

5 CALL INIT
10 FOR I=0 TO 8190 STEP 2 :: CALL PEEK(I,A,B):: TOT=TOT+A*256+B :: NEXT I
20 PRINT TOT

This simply peeks every word in the 8K rom and adds it to TOT.

I will run it but I can tell you now the results will be the same as I am using the same ROM on both machines  

Edited June 18, 2022 by videofx

[senior_falcon]

19 minutes ago, videofx said:

I will run it but I can tell you now the results will be the same as I am using the same ROM on both machines  

I did not know the gram kracker could change roms. If the rom is the same, then the problem must be elsewhere.

I forgot to ask: are you using the F18A?

Edited June 18, 2022 by senior_falcon

[+videofx]

Just now, senior_falcon said:

I did not know the gram kracker could change roms. If the rom is the same, then the problem must be elsewhere

Yup,  The gram Kracker lets you use whatever ROM you want. So I saved the beige console ROM to disk, and then I can use the ROM from my beige TI99 and load it on my Black and silver.

 

Thats why Im stumped if its not the ROM.  The black and silver will run every other version of XB just not 2.9

 

I have other black and silver consoles. Let me see  if I can find another that works and Ill try it. See if its just this particular Ti99.

[senior_falcon]

1 minute ago, videofx said:

Yup,  The gram Kracker lets you use whatever ROM you want. So I saved the beige console ROM to disk, and then I can use the ROM from my beige TI99 and load it on my Black and silver.

 

Thats why Im stumped if its not the ROM.  The black and silver will run every other version of XB just not 2.9

 

I have other black and silver consoles. Let me see  if I can find another that works and Ill try it. See if its just this particular Ti99.

You answered so quickly you might have missed my edit. Are you using the F18A?

[+videofx]

2 minutes ago, senior_falcon said:

You answered so quickly you might have missed my edit. Are you using the F18A?

No F18A, just stock consoles.

[jrhodes]

1 hour ago, senior_falcon said:

That's an interesting observation. I had forgotten that there were some slight differences in the roms.

@videofx

Try this simple program on your two consoles (any XB will work fine) and report the results. This is not super fast, so patience is necessary.

5 CALL INIT
10 FOR I=0 TO 8190 STEP 2 :: CALL PEEK(I,A,B):: TOT=TOT+A*256+B :: NEXT I
20 PRINT TOT

This simply peeks every word in the 8K rom and adds it to TOT.

Classic 99 v399.048 (I have not updated classic 99 in a while), default options, ti xb v110:
75988364

Js99er, default options, ti xb v110:
75988256

Win994a, 3 floppy drives, 32k memory(8bit), ti xb v110:
75981516

I have not tried on my (c)1981 black and silver real iron ti (1mb sams, tipi on sideport, ti xb v110 on fg99)

Edited June 18, 2022 by jrhodes

[senior_falcon]

Looks like there are at least 3 different roms. But since videofx can use the exact operating system on each console, that doesn't seem to be the problem.

@videofx:

I use >83d5 to hold some information about the editing mode. This does not get reset in a "quit" and doesn't seem to be used for anything. If this were somehow modified it could possibly cause trouble. This is a long hail mary pass, but using XB 2.9 G.E.M., can you try this program on the two consoles. Naturally, since you cannot enter anything you would have to use LOAD on the console that fails.

10 CALL PEEK(-31787,X):: PRINT X

 

(edit) I have a feeling this is where the problem lies. I did CALL LOAD(-31787,1) and got 3 extra characters and then the cursor, which was flashing but in a strange way. And it turns out it is unresponsive.

 

 

Edited June 18, 2022 by senior_falcon

[jrhodes]

One more test;

Classic 99 v399.048, ti-99/4 (not A), default options, ti xb v110:
76411323

[senior_falcon]

Let's follow that up with another test. Save this program as LOAD

10 CALL INIT::CALL LOAD(-31787,0)

See if that tames this beast.

[Asmusr]

I think I had to change a few bytes in the ROM in order to implement tape support in JS99er.

[+videofx]

1 hour ago, senior_falcon said:

Let's follow that up with another test. Save this program as LOAD

10 CALL INIT::CALL LOAD(-31787,0)

See if that tames this beast.

@senior_falcon Progress oh guru of the XB

 

I ran the above program and it gives me just a blank blue screen. However, if I power off the PEB, guess what? It dumps me into XB 2.9, then I powered on my PEB and I was able to load and save programs.

 

If I did not run your program and powered off the PEB it did not drop me into XB 2.9

 

 

Edited June 18, 2022 by videofx

[HOME AUTOMATION]

Attempting to clear-up a little perceived confusion here...

5 hours ago, videofx said:

5 hours ago, senior_falcon said:

That's an interesting observation. I had forgotten that there were some slight differences in the roms.

@videofx

Try this simple program on your two consoles (any XB will work fine) and report the results. This is not super fast, so patience is necessary.

5 CALL INIT
10 FOR I=0 TO 8190 STEP 2 :: CALL PEEK(I,A,B):: TOT=TOT+A*256+B :: NEXT I
20 PRINT TOT

This simply peeks every word in the 8K rom and adds it to TOT.

I will run it but I can tell you now the results will be the same as I am using the same ROM on both machines  

And you know this because...

 

5 hours ago, videofx said:

5 hours ago, senior_falcon said:

I did not know the gram kracker could change roms. If the rom is the same, then the problem must be elsewhere

Yup,  The gram Kracker lets you use whatever ROM you want. So I saved the beige console ROM to disk, and then I can use the ROM from my beige TI99 and load it on my Black and silver.

 

Thats why Im stumped if its not the ROM.  The black and silver will run every other version of XB just not 2.9

 

I have other black and silver consoles. Let me see  if I can find another that works and Ill try it. See if its just this particular Ti99.

senior_falcon's, program, sums ROM. GRAMKracker, cannot transpose these.(AFAIK)

Edited June 18, 2022 by HOME AUTOMATION
tranpaose - more commas,,

[Asmusr]

I don't know anything about the GRAM Kracker, but if it could replace the system ROMs then other cartridges could do the same, and I've never heard about that. Imagine being able to put 8K ROM on the 16-bit bus for a game... On the other hand, the system GROMs can be replaced, which the StrangeCart, for instance, can do.

[senior_falcon]

Just goes to show, there is no such thing as unused memory. If I have found it, someone else has beat me to it.

I am going to try an experiment, where that byte is zero'd out at XB startup. It was used to keep the editor status - i.e. 32,40 or 80 column.  So now if you wanted to use the 80 column editor, you would have to type CALL EDIT80. This does not seem like too high a price to pay to make this thing work.

[+videofx]

59 minutes ago, HOME AUTOMATION said:

Attempting to clear-up a little perceived confusion here...

And you know this because...

 

senior_falcon's program, sums ROM. GRAMKracker cannot tranpaose these.(AFAIK)

I know this because as you can see from my photo, this is the ROM that gets loaded on the working and non working TI99. So the checksum would be the same.

[HOME AUTOMATION]

Attempting to relate what I am seeing in your photo, to possible differences in the ROM, is a somewhat disparate process for me, perhaps owing to the TITLE SCREEN's, makeup, residing in GROM, as opposed to ROM.

 

Interestingly, your photo does show that GROM content has been altered from what was likely used during the development of the software currently under testing.

 

GROM\ROM are somewhat independent on the '4a.

[+videofx]

5 minutes ago, HOME AUTOMATION said:

Attempting to relate what I am seeing in your photo, to possible differences in the ROM, is a somewhat disparate process for me, perhaps owing to the TITLE SCREEN's, makeup, residing in GROM, as opposed to ROM.

 

Interestingly, your photo does show that GROM content has been altered from what was likely used during the development of the software currently under testing.

 

GROM\ROM are somewhat independent on the '4a.

The OS resides in GROM 0 , what I have done is taken the OS (grom0) from the working TI99 and via the Gram Kracker I can use that same OS (grom0) on any TI99 as the gram cracker takes control. So we know that the OS has been taken out of the equation. 

 

what ever the issue is it is limited to certain consoles. I went and found another black and silver TI99/4a in my garage and it works.  So not all consoles are the same.

[HOME AUTOMATION]

3 hours ago, videofx said:

So the checksum would be the same.

What I meant to express, is that the sum which senior_falcon, has been referencing, is not that of GROM(s), but that of the ROM, on the address bus, perhaps considered to run the MONITOR(>0000 - >1FFF).

 

I think some '4as, have versions of the 9901, that may cause problems that at first might seem software related.

 

This sounds more like a RAM issue to me, maybe CPU, maybe VDP.

2 hours ago, videofx said:

So not all consoles are the same

Of this, I am certain.

Edited June 18, 2022 by HOME AUTOMATION
spell-un

[+videofx]

I am going to open up both the black and  silver and see if there is anything noticeable.   But @senior_falcon was on the right track. Once I ran his program and turned off the PEB then XB 2.9 worked.

[senior_falcon]

No need for any further analysis. I am 99% sure my fix will work.

[+videofx]

41 minutes ago, senior_falcon said:

No need for any further analysis. I am 99% sure my fix will work.

Did I miss the fix?

[wolhess]

On 6/18/2022 at 5:47 AM, senior_falcon said:

This gets stranger with every post. Let me reiterate what I understand is happening.

You have two consoles, one black & silver and one beige.

Both give a copyright of 1981.

Both return a value of 8 when you CALL PEEKG(256,X)

You have a normal TI expansion box with a gramkracker card.

I assume there is a 32K card in the expansion box. (or does the final grom provide 32K?)

XB 2.9 G.E.M. is on the final grom

The two consoles give different results,with one crashing and one not crashing, even though they are connected exactly the same way and are using the same console groms.

Is the above summary accurate?

It doesn't seem like there should be a difference. If the cartridge is the same, the expansion system is the same, the console groms are the same, the 32K is the same then it is hard to imagine what the difference might be.

Two possibilities come to mind:

Could there be a difference in the 8K console rom?

Is it possible that one of the consoles has 32K built into it?

 

It turns out that CALL PEEKG(258,X)::PRINT X does a better job of differentiating between the systems

TI99/4                        7

TI99/4a (Intern)       190

TI99/4a (Classic99)    80

TI99/4a (V2.2)           45

Hi, I tested the small program "10 CALL PEEKG(258,X)::PRINT X" "20 GOTO 20" in my F18a black&silver console with a 1MB SAMS card in the PEB.

 

In RXB2022D it shows  80, I can break the program and edit is functional

In XB 2.9 it shows        80, I can break the program and the console crashes!

In XB 2.8 it shows        80, I can break the program an edit is functional

In RXB2015 is shows    80, I can break the program an edit is functional

 

[wolhess]

22 hours ago, jrhodes said:

Classic 99 v399.048 (I have not updated classic 99 in a while), default options, ti xb v110:
75988364

Js99er, default options, ti xb v110:
75988256

Win994a, 3 floppy drives, 32k memory(8bit), ti xb v110:
75981516

I have not tried on my (c)1981 black and silver real iron ti (1mb sams, tipi on sideport, ti xb v110 on fg99)

Hi,

I run the program from post #25

The console is my black&silver (German PAL version) with F18a, PEB with 1MB SAMS, TI Disk Controller, 3 Floppy drives, tipipeb card

RXB 2015, RXB 2022D, XB 2.8, XB 2.9

The result is in every XB the same:  75988364

 

[wolhess]

22 hours ago, senior_falcon said:

Looks like there are at least 3 different roms. But since videofx can use the exact operating system on each console, that doesn't seem to be the problem.

@videofx:

I use >83d5 to hold some information about the editing mode. This does not get reset in a "quit" and doesn't seem to be used for anything. If this were somehow modified it could possibly cause trouble. This is a long hail mary pass, but using XB 2.9 G.E.M., can you try this program on the two consoles. Naturally, since you cannot enter anything you would have to use LOAD on the console that fails.

10 CALL PEEK(-31787,X):: PRINT X

 

(edit) I have a feeling this is where the problem lies. I did CALL LOAD(-31787,1) and got 3 extra characters and then the cursor, which was flashing but in a strange way. And it turns out it is unresponsive.

 

 

In XB2.8 I did the CALL PEEK(-31787,x):: PRINT X and got "14"; A CALL LOAD(-31787,1) works fine; another CALL PEEK(-31787,x):: PRINT X results in the expected "1".

I did this in XB 2.9 "10 CALL PEEK(-31787,x):: PRINT X"  "20 GOTO 20" and got "0". Then I could break the program without an issue to the prompt. Now I can edit in XB in a normal way.

 

Next I powercycled the console

 

And I chosed XB 2.9 from the FG99 menu

I saw the number "14" again; A break shows me the three extra characters and the flickering cursor. The console was crashed!

 

 

 

[+mizapf]

19 hours ago, Asmusr said:

I don't know anything about the GRAM Kracker, but if it could replace the system ROMs then other cartridges could do the same, and I've never heard about that. Imagine being able to put 8K ROM on the 16-bit bus for a game... On the other hand, the system GROMs can be replaced, which the StrangeCart, for instance, can do.

It can only override the GROMs, not the system ROMs. The GRAM Kracker used CMOS SRAMs, battery-backed.

 

gram kracker manual.pdf