---

----------------------------------------

X - PRG 0 VS PRG 1

----------------------------------------



Now we have resolved the mystery about the Fighter's Ring, and that the

Death Necklace can only be acquired once per game. But we have done these

examinations with one version of the ROM. Couldn't it be possible that there

is some other version out there that is different with these regards?



Surely, if we remember getting more than one Death Necklace per game, and we

proved that this is not possible, there must be some other version out there

with this difference, right? If so, maybe this other version has the

Fighter's Ring's issue fixed as well? Pretty please? With an opcode on top?



Throughout this FAQ, I have looked at the PRG 0 version of the game. The

issue with the Death Necklace got me thinking about this. If we do a diff of

the PRG 0 and PRG 1 versions of the game using some utility out there that

does this, we see that the PRG 0 and PRG 1 versions are different by exactly

3 bytes:



Address     PRG 0     PRG 1

3FAE        37        32

3FAF        32        29

AF7C        EF        F0



The first 2 addresses have to do with the title screen.



In PRG 0, the title screen says at the bottom:



LICENSED TO NINTENDO

TM TRADEMARK TO NINTENDO



In PRG 1, the title screen says at the bottom:



LICENSED TO NINTENDO

TM TRADEMARK OF NINTENDO



So basically, 66 % of the differences between PRG 0 and PRG 1 was to correct

the grammar... at the title screen where everyone can see it openly and poke

fun of it...



This is one easy way to tell which version of the game you have on the cart,

assuming there are only 2 versions, which, according to our research on the

Death Necklace, does not seem to be the case at all...



The last address seems like it can be code. If we do a seek to the relevant

address on the game, taking into account the NES header, we see the

following:



PRG 0:



$AF6B:1D EF 5F  ORA $5FEF,X

$AF6E:0D 0E 0C  ORA $0C0E

$AF71:1B        UNDEFINED



PRG 1:



$AF6B:1D F0 5F  ORA $5FF0,X

$AF6E:0D 0E 0C  ORA $0C0E

$AF71:1B        UNDEFINED



Before we even waste our time analyzing these opcodes, note that after we do

our OR operations, we eventually get to an undefined opcode. If these were

truly supposed to be opcodes, we can see that clearly, their execution would

lead to an undefined opcode before any branching could occur, which would

crash our game. Therefore, these bytes couldn't logically be opcodes. They

are probably something else. Even if these were valid opcodes, this single

difference here does not address any of the relevant details we have

researched. We can conclude that, for purposes of this FAQ, these two

versions are identical.





UPDATE 1.1:





Going back to this matter, one way we can determine what this last byte is

used for is via ROM corruption. Doing so and playing the game for a while to

verify the effect, we find out that the last byte actually has to do with

text as well.



If we look at a chunk of data through a hex editor that has dictionary

capabilities, and we use the game's text as the dictionary, we see the

following:



PRG 0:



AF60    0E 19 47 FC 60 FD 37 11    ep.. .Th

AF68    0E 5F F4 5F 0A 1D 1D 0A    e . atta

AF70    0C 14 1C 4C FC 37 11 22    cks!.Thy

AF78    5F 2B 12 1D EF 5F 0D 0E     Hit. de

AF80    0C 1B 0E 0A 1C 0E 0D 5F    creased 

AF88    0B 22 5F F5 47 FC 24 5F    by ...A 



PRG 1:



AF60    0E 19 47 FC 60 FD 37 11    ep.. .Th

AF68    0E 5F F4 5F 0A 1D 1D 0A    e . atta

AF70    0C 14 1C 4C FC 37 11 22    cks!.Thy

AF78    5F 2B 12 1D F0 5F 0D 0E     Hit. de

AF80    0C 1B 0E 0A 1C 0E 0D 5F    creased 

AF88    0B 22 5F F5 47 FC 24 5F    by ...A 



This clearly corresponds to the text we see when we get attacked by an enemy

in battle. Getting into a battle, we see the following:



PRG 0, HP Damage = 1:



The Slime attacks!

Thy Hit decreased by 1.



PRG 1, HP Damage = 1:



The Slime attacks!

Thy Hit Point decreased by 1.



PRG 0, HP Damage > 1:



The Slime attacks!

Thy Hits decreased by 2.



PRG 1, HP Damage > 1:



The Slime attacks!

Thy Hit Points decreased by 2.



So we see that the value of F0 corresponds to a single byte representation

of the string " Point", and the value of EF, when interpreted as such,

simply prints out null, or nothing at all. This kind of text "compression"

is actually quite common among video games.



From this, we can see that all 3 bytes that are different between PRG 0 and

PRG 1 correspond to text differences in the game. This means that PRG 0 and

PRG 1 SRAM data, save states, movie files, Game Genie codes, and others are

perfectly backwards-compatible with each other.



Also, it seems that I need to correct myself: 100 % of the differences

between PRG 0 and PRG 1 was to correct the grammar... It's nice to finally

know what the real differences are between the two versions! :)





/UPDATE

---