Glitch City Laboratories Archives

Sadly I don't know if they have hosted their code :(

Sadly I don't know if they have hosted their code :(

Good news it appears offgao has hosted the Poketetris and what is probably the memory editor codes on his website http://offgao.net Specifically on the section about Pokemon http://offgao.net/poke/
It appears that he/she is using bytes for box storage for tetris and nicknames for the editor? (I used an English R/B RAM map and am unsure of how fare off JP is.)
It also seems that he/she thinks Poketetris is impossible in English R/B? "海は" which google translate says means "Overseas version is probably impossible." However I know nothing of Japanese and google translate is already known to not be a 100% accurate source. Hopefully the Memory editor can be used overseas as that seems like one of the most useful innovations in years. Although it might need to be rewritten since it seems that party Pokemon are corrupted to an extent?
Also if anyone can read Japanese the video with the memory editor appears to have another fun looking game in it that he/she might explain in it.

Given that box data starts at 0xDE06 in the English versions of Pokemon Red/Blue, it seems like this could totally be possible, especially by using 8F. Unfortunately, it doesn't seem like it'd persist between resets as-is since the last 3 bytes would be truncated to fit the 440-byte limit. (the code is 446 bytes)

Storing it elsewhere (some parts of WRAM) or modifying the code a bit to make it fit should make it persist across resets, but it definitely does work after a few modifications, such as changing the call $3867 instruction to call $3831, which is the JoypadLowSensitivity function in the English version.

[img]http://i.imgur.com/naRbemp.png[/img]

I've attached the source code with modifications to make it work with the English versions when executed from $DE21.

Given that box data starts at 0xDE06 in the English versions of Pokemon Red/Blue, it seems like this could totally be possible, especially by using 8F. Unfortunately, it doesn't seem like it'd persist between resets as-is since the last 3 bytes would be truncated to fit the 440-byte limit. (the code is 446 bytes)

Given that box data starts at 0xDE06 in the English versions of Pokemon Red/Blue, it seems like this could totally be possible, especially by using 8F. Unfortunately, it doesn't seem like it'd persist between resets as-is since the last 3 bytes would be truncated to fit the 440-byte limit. (the code is 446 bytes)

Unless Datacrystal's RAM map (found here http://datacrystal.romhacking.net/wiki/Pok%C3%A9mon_Red/Blue:RAM_map#Stored_Pok.C3.A9mon) is off box data starts at DA96. According to Datacrystal DE06 is the start of Pokemon nicknames which would mean that box data is definitely large enough space seeing as according to the same source box data ends at DEE1 which would give you 1099 Bytes to work with.

It seems hl becomes DEFD, so I tried placing the graphics there in case that is where the game expects to find the graphics nd the graphics changed but they were still loaded wrongly.

Do you know what changes I need to make to my code to make it work properly? Thanks!

…box data starts at DA96. According to Datacrystal DE06 is the start of Pokemon nicknames which would mean that box data is definitely large enough space seeing as according to the same source box data ends at DEE1 which would give you 1099 Bytes to work with.

It seems hl becomes DEFD, so I tried placing the graphics there in case that is where the game expects to find the graphics nd the graphics changed but they were still loaded wrongly.

Do you know what changes I need to make to my code to make it work properly? Thanks!

Sure thing! To make it work when executing from $D321, the following modification will do the trick:

Line 125:
[tt]ld h,$DE  =>  ld h,$D3[/tt]

Also, the final [tt]db $65[/tt] instruction must end on byte $xxFF because of how the code refers to the necessary sprites. I'll be working on shortening the code today so that it can be done using PC boxes in the English version of R/B.

Ah, cool. Will this work if D321 contains jp DA80 too or would more adjustments have to be made?

Ah, cool. Will this work if D321 contains jp DA80 too or would more adjustments have to be made?

Nah, it'd only work if that code was at D321 and was executed from there. That's a good point though, I'll modify it so that it's not entirely location-dependent. (e.g. doesn't need to end on an xxFF boundary)

I see. Thanks! Yeah, it would be a good idea to fit a version in DA80 (the actual start of the boxed data) due to map data changing for every step, etc. There is enough room because current box data is a whopping 1122 bytes.

I see. Thanks! Yeah, it would be a good idea to fit a version in DA80 (the actual start of the boxed data) due to map data changing for every step, etc. There is enough room because current box data is a whopping 1122 bytes.

Okay, so I came up with a method that uses the box, but it involves glitch Pokemon and normally-unobtainable values.

Using 8F to insert the Pokemon into the box is totally worth it though, since it's essentially a permanent hex editor.

Only 7 Pokemon are used; just insert this at DA80 and you'll be good to go. ;D


07F60100FE01181FFF00000000000000000000000000
F6F7F8F9FAFBFCFDFEFF606162636465660100DAC50B0B161221AAC3CD4FDB03C5
010A0009C11520F3C13EEDEAD3C33EFFE0B776C5CD3138C1F0B5FE02C8FE032003
00C5C9FE4020010BFE80200103FE20200621F0FF09444DFE10200621100009444D
FE2420062100FF09444DFE14200621000109444D1600FE88200216FFFE48200216
01FE28200216F0FE18200216106069FE442008F53AF52A32F177F1FE84200AF52A
1800F53A22F177F1444D7AA728030A8202C3AADA3E7C223E7F2278CD6ADB791E1F
CD6ADB3EE3220ACD6ADB3E7C22C9F5CB37CD71DBF1E526DAE60FC6966F7EE122C9


[img]http://i.imgur.com/nYSuPjI.png[/img]

I've also attached the ASM equivalent of the code above.