Nukey's Pacman 8K on Flashback

[dr.geek]

I tried playing Nukey's Pacman 8K on my Flashback 9 via SD Card and it refuses to work (I get a black screen with some random graphics at the top of the screen and a buzzing noise) - I have read where people have been playing it on FB9...My FB9 is an Australian model. Is there a differenece between the Australian and USA models?

 

Also having trouble with some homebrews (HALO 2600 for example)

Edited September 18, 2019 by dr.geek

[Nukey Shay]

That's Dintar816's homebrew that people mentioned.

 

Flashback Portable refuses to run my hack, I never discovered why (it's not doing anything special).  I gather that FB9 is the same.

 

[Mr SQL]

On 9/21/2019 at 3:23 AM, Nukey Shay said:

That's Dintar816's homebrew that people mentioned.

 

Flashback Portable refuses to run my hack, I never discovered why (it's not doing anything special).  I gather that FB9 is the same.

 

If your 8K hack isn't using enough of the second bank it won't run, if this is the case you could fill some of the extra space with data so the Portable will recognize the bank. 

[Nukey Shay]

Both banks are nearly full...much more than Star Raiders is (which is a game that I got working for the AFP).  F8 standard banking only in both cases, no added Ram or whatnot. AFAIK, the flashback emulator is the only platform Pac-Man8k fails on.

[PiLHA]

On 11/3/2019 at 4:15 PM, Mr SQL said:

If your 8K hack isn't using enough of the second bank it won't run, if this is the case you could fill some of the extra space with data so the Portable will recognize the bank. 

Does this rule apply to any bank? Should any of them have use to be recognized? How many KB is this possible?

[Nukey Shay]

The one rule is that there are no rules.  What works in one case might not in another.  Consider playfield register updates: at the latest, these must occur at specific cycle counts.  Most of the time the portable follows spec.  Other times, the cycle count needed will deviate for no discernible reason and need a bit of trial/error to fix.  CTRLPF is a real mind bender...often when called midscanline (or even at the end of a scanline), it will affect everything that had already been drawn.  And of course, VBLANK...no rhyme or reason why it pushes the display in some programs but not in others.

[Mr SQL]

On 11/4/2019 at 7:26 PM, PiLHA said:

Does this rule apply to any bank? Should any of them have use to be recognized? How many KB is this possible?

Yes it generally only affects homebrews and hacks since the commercial releases usually filled the banks, interesting detection method.

 

Chapter 17 has the details on how to add padding to get partially used banks to be recognized, Assembly or batari BASIC programs that are similarly affected can be fixed the same way:

http://relationalframework.com/Atari_Flashback_BASIC_Manual.txt

 

 

[PiLHA]

3 hours ago, Mr SQL said:

Yes it generally only affects homebrews and hacks since the commercial releases usually filled the banks, interesting detection method.

 

Chapter 17 has the details on how to add padding to get partially used banks to be recognized, Assembly or batari BASIC programs that are similarly affected can be fixed the same way:

http://relationalframework.com/Atari_Flashback_BASIC_Manual.txt

 

 

Thank you. Finally figured out how to do it.

[Nukey Shay]

Pacman8k hack fixed!

https://atariage.com/forums/applications/core/interface/file/attachment.php?id=685915

[+Karl G]

What was needed to get it to work? 

 

Edit: nevermind; I see the answer in the other topic. 

[Nukey Shay]

For any future necrobumps...

I believe that the portable's emulator treated this game as using Superchip Ram - effectively making all the bankswitch instructions and all other program code on the lowest page of Rom unable to execute.  I already had enough free space in both banks to keep the lowest page unused, so even if it still misreads it as a Superchip program, it doesn't matter.