Thread Status:
Not open for further replies.
  1. Kroatien99

    OP Kroatien99 GBAtemp Regular
    Member

    Joined:
    Mar 14, 2010
    Messages:
    140
    Country:
    Croatia
    Please Post just ideas how to solve problems. <img src="style_emoticons/<#EMO_DIR#>/yaywii.gif" style="vertical-align:middle" emoid=":yaywii:" border="0" alt="yaywii.gif" />
    - It works 100% on AKAIO RC 1.6
    - Tried to apply the JP fix : Fail
    - Tried to move US files to JP one, to use that JP fix : Epic Fail
    - Tried to patch ARM7 with JP HG/SS one, Platinum one, D/P one : Fail
    - Tried to patch ARM9 with JP HG/SS one, Platinum one, D/P one : Fail
    - Defined "events" as "when you enter/exit a building/stairs/door or save or encounter a wild pokemon or start a trainer fight or end a battle (get back to the game) or when you end a cutscene"
    - The number of events "needed" before the game freezes depend of the number of badges you have
    - Replacing all the 00 00 00 46 in the rom by 00 00 00 00 : Fail
    - Tried a lot of AR codes : Some from JP, K fix... No one working
    - Unchecking DMA and reset : Fail
    - Inserting patched JP overlays : Fail, but increase the number of events before freeze
    - Deleting overlays : Fail
    - Dissect the AKAIO firmware and find out how theirs works: Fail-encrypted
    - You replace x with the number of badges you have, and you find : 0 badges 24 events, 1 badges 21 events, 2 badges 15 event, 3 badges 3 events, more badges you have a negative value and the game do not want to load the save
    - With this "events" needed we can think there is a countdown, each event add -1 to a var who doesn't support value below 1, that can explain the countdown crashes and the fact that the game won't start if the value is negative. If it's right, a AR code who lock
    - No$gba:
    Pokemon HeartGold (U) black screen fix code:
    020DD9E4 E1A00000
    020D3820 E1A00000

    Pokemon Soul Silver (U) black screen fix code:
    020DE16C E1A00000
    020D3FA8 E1A00000
    This list was made by Ribesg

    <!--coloro:#000000--><span style="color:#000000"><!--/coloro--><!--sizeo:5--><span style="font-size:18pt;line-height:100%"><!--/sizeo-->Information from Ribesg added by Hadriano<!--sizec--></span><!--/sizec--><!--colorc--></span><!--/colorc-->


    <!--sizeo:5--><span style="font-size:18pt;line-height:100%"><!--/sizeo--><b>THE RECAP POST - UPDATE</b><!--sizec--></span><!--/sizec-->

    What was tried, what we know :
    - It works <b>95%</b> on <b>AKAIO RC 1.6</b> (Some crash report, but works really better than on other flashcards)
    - Tried to apply the JP fix : <b>Fail</b>
    - Tried to move US files to JP one, to use that JP fix : <b>Epic Fail</b>
    - <b>Data folder</b> from US was moved to the patched JP rom : <b><i>Strange Result</i></b>
    - Tried to patch <b>ARM7</b> with JP HG/SS one, Platinum one, D/P one : <b>Fail</b>
    - Tried to patch <b>ARM9</b> with JP HG/SS one, Platinum one, D/P one : <b>Fail</b>
    - Defined "<b>events</b>" as "when you enter/exit a building/stairs/door or save or encounter a wild pokemon or start a trainer fight or end a battle (get back to the game) or when you end a cutscene"
    - The number of events "needed" before the game freezes <b>depend of the number of badges</b> you have
    - Replacing all the <i>00 00 00 46</i> in the rom by <i>00 00 00 00</i> : <b>Fail</b>
    - Tried <b><u>a lot of AR codes</u></b> : Some from JP, K fix... No one working
    - Everytime the game crashes there is an access violation of dma at offset <b>56612</b>
    - Unchecking DMA and reset : <b>Fail</b>
    - Inserting patched JP overlays : <b>Fail</b>, but increase the number of events before freeze
    - Deleting overlays : <b>Fail</b>
    - This match with the equation <b>y = -(1/2)*x^3 - (5/2)*x + 24</b> and with <b>y = 27 - 3*(2^x)</b> and certainly with some other ones.
    - You replace x with the number of badges you have, and you find : 0 badges 24 events, 1 badges 21 events, 2 badges 15 event, 3 badges 3 events, more badges you have a negative value and the game do not want to load the save
    - With this "events" needed we can think there is a countdown, each event add -1 to a var who doesn't support value below 1, that can explain the countdown crashes and the fact that the game won't start if the value is negative. If it's right, a AR code who lock this var must work...
    Some of what can be done :
    - Try to know why it works on AKAIO 1.6 RC1, we need some dev of this firware to help us understand
    - Try to find where this equation is stored, if it exists
    - Try to find where the "countdown" is stored, it appears that they are multiple var
    - Try to Hex read the ROM and/or the files in the ROM in order to find new interesting things
    - Ask to yourself : why it works with official games, and not on our flashcards ?
    What can be usefull (Or not) :
    By hexing, I found out some nice stuff:

    [SDK+NINTENDO:DWC2.2.30008.080630.1906_DWC_2_2_PLUS8]
    [SDK+NINTENDO:BACKUP]
    [SDK+Abiosso:libVCT 1.3.1]
    [SDK+NINTENDO:WiFi2.1.30003.0709200229]
    [SDK+UBIQUITOUS:CPS]
    [SDK+UBIQUITOUS:SSL]

    SSL and CPS, isn't that the encryption systems? Also, I think this Abiosso is the AP software... not sure.
    YOOOOOO, I FOUND SOMETHING NEW, remember this:

    [SDK+NINTENDO:DWC2.2.30008.080630.1906_DWC_2_2_PLUS8]

    NOW, there is a folder called DWC!!!! AND in that folder, the utility.bin file gives this:
    <!--quoteo--><div class='quotetop'>QUOTE</div><div class='quotemain'><!--quotec-->eb2Ap.nsc.l
    eb2HlAp.nsc.leb2HlWiFi.nsc.l
    eb2Menu.nsc.leb3HlList1.nsc.leb3HlList2.nsc.leb3HlList3.nsc.l
    eb3List.nsc.leb3ListBack.nsc.l eb3Way.nsc.leb4ApList.nsc.leb4Error.nsc.leb4HlDns0.nsc.leb4HlDns1.nsc.leb4H
    Gateway.nsc.l
    eb4HlIp.nsc.leb4HlMask.nsc.leb4HlSet1.nsc.leb4HlSet2.nsc.leb4HlSet3.nsc.leb
    HlSsid.nsc.leb4HlUsb.nsc.leb4HlWep.nsc.l eb4Usb.nsc.leb5HlErase.nsc.leb5HlInfo.nsc.leb5HlMove.nsc.leb5HlOption.nsc.l
    eb5Info.nsc.l
    eb5Move.nsc.leb5OptMenu.nsc.l ebBgHl.ncg.lebBgOption.ncg.lebBgStep1.ncg.lebBgStep1.ncl.lebBgStep2.ncg.leb
    gStep21.ncg.lebBgStep3.ncg.l ebMain.nce.lebObjMain.ncg.letBgMain.ncg.letBgMain.ncl.l etMain.nce.l etNull.nsc.letObjMain.ncg.letOption.nsc.l
    etStep1.nsc.l
    etStep2.nsc.l
    etStep3.nsc.l
    etTop.nsc.l
    fb2Ap.nsc.l
    fb2HlAp.nsc.lfb2HlWiFi.nsc.l
    fb2Menu.nsc.lfb3HlList1.nsc.lfb3HlList2.nsc.lfb3HlList3.nsc.l
    fb3List.nsc.lfb3ListBack.nsc.l fb3Way.nsc.lfb4ApList.nsc.lfb4Error.nsc.lfb4HlDns0.nsc.lfb4HlDns1.nsc.lfb4H
    Gateway.nsc.l
    fb4HlIp.nsc.lfb4HlMask.nsc.lfb4HlSet1.nsc.lfb4HlSet2.nsc.lfb4HlSet3.nsc.lfb
    HlSsid.nsc.lfb4HlUsb.nsc.lfb4HlWep.nsc.l fb4Usb.nsc.lfb5HlErase.nsc.lfb5HlInfo.nsc.lfb5HlMove.nsc.lfb5HlOption.nsc.l
    fb5Info.nsc.l
    fb5Move.nsc.lfb5OptMenu.nsc.l fbBgHl.ncg.lfbBgOption.ncg.lfbBgStep1.ncg.lfbBgStep1.ncl.lfbBgStep2.ncg.lfb
    gStep21.ncg.lfbBgStep3.ncg.l fbMain.nce.lfbObjMain.ncg.lftBgMain.ncg.lftBgMain.ncl.l ftMain.nce.l ftNull.nsc.lftObjMain.ncg.lftOption.nsc.l
    ftStep1.nsc.l
    ftStep2.nsc.l
    ftStep3.nsc.l
    ftTop.nsc.l
    gb2Ap.nsc.l
    gb2HlAp.nsc.lgb2HlWiFi.nsc.l
    gb2Menu.nsc.lgb3HlList1.nsc.lgb3HlList2.nsc.lgb3HlList3.nsc.l
    gb3List.nsc.lgb3ListBack.nsc.l gb3Way.nsc.lgb4ApList.nsc.lgb4Error.nsc.lgb4HlDns0.nsc.lgb4HlDns1.nsc.lgb4H
    Gateway.nsc.l
    gb4HlIp.nsc.lgb4HlMask.nsc.lgb4HlSet1.nsc.lgb4HlSet2.nsc.lgb4HlSet3.nsc.lgb
    HlSsid.nsc.lgb4HlUsb.nsc.lgb4HlWep.nsc.l gb4Usb.nsc.lgb5HlErase.nsc.lgb5HlInfo.nsc.lgb5HlMove.nsc.lgb5HlOption.nsc.l
    gb5Info.nsc.l
    gb5Move.nsc.lgb5OptMenu.nsc.l gbBgHl.ncg.lgbBgOption.ncg.lgbBgStep1.ncg.lgbBgStep1.ncl.lgbBgStep2.ncg.lgb
    gStep21.ncg.lgbBgStep3.ncg.l gbMain.nce.lgbObjMain.ncg.lgtBgMain.ncg.lgtBgMain.ncl.l gtMain.nce.l gtNull.nsc.lgtObjMain.ncg.lgtOption.nsc.l
    gtStep1.nsc.l
    gtStep2.nsc.l
    gtStep3.nsc.l
    gtTop.nsc.l
    ib2Ap.nsc.l
    ib2HlAp.nsc.lib2HlWiFi.nsc.l
    ib2Menu.nsc.lib3HlList1.nsc.lib3HlList2.nsc.lib3HlList3.nsc.l
    ib3List.nsc.lib3ListBack.nsc.l ib3Way.nsc.lib4ApList.nsc.lib4Error.nsc.lib4HlDns0.nsc.lib4HlDns1.nsc.lib4H
    Gateway.nsc.l
    ib4HlIp.nsc.lib4HlMask.nsc.lib4HlSet1.nsc.lib4HlSet2.nsc.lib4HlSet3.nsc.lib
    HlSsid.nsc.lib4HlUsb.nsc.lib4HlWep.nsc.l ib4Usb.nsc.lib5HlErase.nsc.lib5HlInfo.nsc.lib5HlMove.nsc.lib5HlOption.nsc.l
    ib5Info.nsc.l
    ib5Move.nsc.lib5OptMenu.nsc.l ibBgHl.ncg.libBgOption.ncg.libBgStep1.ncg.libBgStep1.ncl.libBgStep2.ncg.lib
    gStep21.ncg.libBgStep3.ncg.l ibMain.nce.libObjMain.ncg.litBgMain.ncg.litBgMain.ncl.l itMain.nce.l itNull.nsc.litObjMain.ncg.litOption.nsc.l
    itStep1.nsc.l
    itStep2.nsc.l
    itStep3.nsc.l
    itTop.nsc.l
    jb2Ap.nsc.l
    jb2HlAp.nsc.ljb2HlWiFi.nsc.l
    jb2Menu.nsc.ljb3HlList1.nsc.ljb3HlList2.nsc.ljb3HlList3.nsc.l
    jb3List.nsc.ljb3ListBack.nsc.l jb3Way.nsc.ljb4ApList.nsc.ljb4Error.nsc.ljb4HlDns0.nsc.ljb4HlDns1.nsc.ljb4H
    Gateway.nsc.l
    jb4HlIp.nsc.ljb4HlMask.nsc.ljb4HlSet1.nsc.ljb4HlSet2.nsc.ljb4HlSet3.nsc.ljb
    HlSsid.nsc.ljb4HlUsb.nsc.ljb4HlWep.nsc.l jb4Usb.nsc.ljb5HlErase.nsc.ljb5HlInfo.nsc.ljb5HlMove.nsc.ljb5HlOption.nsc.l
    jb5Info.nsc.l
    jb5Move.nsc.ljb5Multi.nsc.ljb5OptMenu.nsc.l jbBgHl.ncg.ljbBgOption.ncg.ljbBgOption.ncl.ljbBgOption1.ncl.ljbBgStep1.ncg.l
    jbBgStep1.ncl.ljbBgStep11.ncl.ljbBgStep2.ncg.ljbBgStep2.ncl.ljbBgStep21.ncg.
    jbBgStep21.ncl.ljbBgStep3.ncg.ljbBgStep3.ncl.ljbBgStep31.ncl.l jbMain.nce.l
    jbObjKb.ncl.ljbObjMain.ncg.ljbObjMain.ncl.ljbObjWay.ncl.ljtBgMain.ncg.ljtBg
    ain.ncl.l jtMain.nce.l jtNull.nsc.ljtObjMain.ncg.ljtOption.nsc.l
    jtStep1.nsc.l
    jtStep2.nsc.l
    jtStep3.nsc.l
    jtTop.nsc.l
    kb2Ap.nsc.l
    kb2HlAp.nsc.lkb2HlWiFi.nsc.l
    kb2Menu.nsc.lkb3HlList1.nsc.lkb3HlList2.nsc.lkb3HlList3.nsc.l
    kb3List.nsc.lkb3ListBack.nsc.l kb3Way.nsc.lkb4ApList.nsc.lkb4Error.nsc.lkb4HlDns0.nsc.lkb4HlDns1.nsc.lkb4H
    Gateway.nsc.l
    kb4HlIp.nsc.lkb4HlMask.nsc.lkb4HlSet1.nsc.lkb4HlSet2.nsc.lkb4HlSet3.nsc.lkb
    HlSsid.nsc.lkb4HlUsb.nsc.lkb4HlWep.nsc.l kb4Usb.nsc.lkb5HlErase.nsc.lkb5HlInfo.nsc.lkb5HlMove.nsc.lkb5HlOption.nsc.l
    kb5Info.nsc.l
    kb5Move.nsc.lkb5OptMenu.nsc.l kbBgHl.ncg.lkbBgOption.ncg.lkbBgStep1.ncg.lkbBgStep1.ncl.lkbBgStep2.ncg.lkb
    gStep21.ncg.lkbBgStep3.ncg.l kbMain.nce.lkbObjMain.ncg.lktBgMain.ncg.lktBgMain.ncl.l ktMain.nce.l ktNull.nsc.lktObjMain.ncg.lktOption.nsc.l
    ktStep1.nsc.l
    ktStep2.nsc.l
    ktStep3.nsc.l
    ktTop.nsc.l
    sb2Ap.nsc.l
    sb2HlAp.nsc.lsb2HlWiFi.nsc.l
    sb2Menu.nsc.lsb3HlList1.nsc.lsb3HlList2.nsc.lsb3HlList3.nsc.l
    sb3List.nsc.lsb3ListBack.nsc.l sb3Way.nsc.lsb4ApList.nsc.lsb4Error.nsc.lsb4HlDns0.nsc.lsb4HlDns1.nsc.lsb4H
    Gateway.nsc.l
    sb4HlIp.nsc.lsb4HlMask.nsc.lsb4HlSet1.nsc.lsb4HlSet2.nsc.lsb4HlSet3.nsc.lsb
    HlSsid.nsc.lsb4HlUsb.nsc.lsb4HlWep.nsc.l sb4Usb.nsc.lsb5HlErase.nsc.lsb5HlInfo.nsc.lsb5HlMove.nsc.lsb5HlOption.nsc.l
    sb5Info.nsc.l
    sb5Move.nsc.lsb5OptMenu.nsc.l sbBgHl.ncg.lsbBgOption.ncg.lsbBgStep1.ncg.lsbBgStep1.ncl.lsbBgStep2.ncg.lsb
    gStep21.ncg.lsbBgStep3.ncg.l sbMain.nce.lsbObjMain.ncg.lstBgMain.ncg.lstBgMain.ncl.l stMain.nce.l stNull.nsc.lstObjMain.ncg.lstOption.nsc.l
    stStep1.nsc.l
    stStep2.nsc.l
    stStep3.nsc.l
    stTop.nsc.lxb3Multi.nsc.lxb4ApListBack.nsc.l
    xb4Edit.nsc.lxb4EditAddr.nsc.lxb4Multi.nsc.l
    xb4None.nsc.lxtObjMain.ncl.lyb5Multi.nsc.lybBgOption.ncl.lybBgOption1.ncl.l
    bBgStep11.ncl.lybBgStep2.ncl.lybBgStep21.ncl.lybBgStep3.ncl.lybBgStep31.ncl.

    ybObjKb.ncl.lybObjMain.ncl.lybObjWay.ncl.l
    banner.char
    banner.plt child.srl eng.bmg.l fre.bmg.l ger.bmg.l ita.bmg.l jap.bmg.l
    kc_m.NFTR.l kor.bmg.l
    lc_m.NFTR.l
    lc_s.NFTR.l spa.bmg.l usa.bmg.l sound_data.sdat.l<!--QuoteEnd--></div><!--QuoteEEnd-->
    This person claims the AP is a file called Guru2.

    This is basically the same no$gba fix as before, but he give the steps as to how he did it.

    What this did
    I removed the AP from the ROM
    This means there is no longer Black screens
    Guys, guru2.narc is useless, I analyzed and it is related to graphic data: panels, polygons, egg01, egg02, pSphere, material, etc...

    Yeah. I only found things about polygons and stuff too, but didn't someone say that after a while all the tiles went black, and then another guy claimed it had something to do with the AP? So...who knows...
    OK, here is some help from Normmatt, who stated that there are multiple checks.
    Firstly, Rudolph's Patcher apparently fixes the first check.

    020DE16C E1A00000
    This fixes the second check, but is not done correctly - only disables it. It should fool it, not disable it.

    Basically, what we have to do is to find the address rudolph's patcher changes, and patch the ROM with both that change and a better version of the code above.
    Those were the difference on Patched and non patched J version.
    <img src="http://img194.imageshack.us/img194/9499/20100314130113.png" border="0" class="linked-image" /><img src="http://img123.imageshack.us/img123/9566/20100314130856.png" border="0" class="linked-image" />
    I HAVE READ THESE POSTS AND HAVE A SOLUTION, ALMOST

    I can't do it, but I know the logic behind it.

    I remember many posts back some guy posted a peice of code found, that changed per event; this obviously is it!

    HOWEVER, as the person had said, removing said code via ar, doesnt work since the game needs it.

    SO WHAT DO? SIMPLE!

    Remove the code in ar, then place the code, every time, so the count stays at 24, virtually fixing all freezing issues.


    I need to find the post in the other thread, but if it is possible to insert code via ar, this can be fixed!

    Yes, I know my first post here, blah blah..
    once again earlier we did find 2 possible addresses
    the first was 02260344 this one is constantly rewritten when any event takes place
    the second is at 04000600 and it 2 is rewritten on screen loading with the values of
    02080086
    00000086
    00000046-commanality
    these were looked at but freezes still took place give me a min i think i just got a new code 2 try
    Yo guys, according to this: <a href="http://www.pokecommunity.com/showthread.php?t=194248" target="_blank">http://www.pokecommunity.com/showthread.php?t=194248</a>
    The guy's personal findings list includes this:

    "The pokegra.narc in the pbr folder uses the D/P encryption"???

    HeartGold and SoulSilver have that file in the PBR folder too....
    I analyzed the narc file and it contains around 3000 files in it...
    a couple of possible locations in the utilities.bin file
    !r"r#r$r%r&r'r®r*r+r,r-r.r
    found at offset 15520-30
    @ B!B"B#B$B%B&B'B(B)B*B+B,B-B.B/B0B1B2B3B4B
    offset 174f0-17510
    both of these seems to once again use math to do something i'm not sure what
    I just needed to have this with: <!--quoteo--><div class='quotetop'>QUOTE</div><div class='quotemain'><!--quotec-->Hello to all I am French, and I have to find madman's thing, I have patcher the version ( FR) with a patch of tradu FR in the bass for THE VERSION ( J ) but patch in controuner the protection games plants more nor of frezzz

    Then I believes in patch it has the solution of the protection of games I give you the link if wanted him(it) to you

    I am strong in no way there THERE I am to sadden

    BUT THE PATCH WALKS<!--QuoteEnd--></div><!--QuoteEEnd-->
    Hmm, I'm still watching some RAM addresses with DeSmuME.

    You guys remember the infamous 02260344 address?
    I've found 4 addresses more that have the exact same behaviour, they all seem to do some calculations when an event occurs.
    021E756D
    021F1017
    021F1083
    021F1205

    these address cant be skipped like the others can they are more then like something to do with general screen change
    when given the skip code it remains white
    gimme a lil while to tweak with them i might b able to find 1 that works individually
    that line of code appears multiple times in the utility bin
    ultility bin
    offset ddc40 of
    €9MESGbmg1€90[email protected]
    offset db3b0-c0
    ÀMESGbmg1À0[email protected]
    offset d8ec0-e2
    ÿÿÿMESGbmg10[email protected]
    offset d801c-35
    MESGbmg1€90INF1
    offset 0d7080-a1
    MESGbmg1`:0[email protected]
    offset d6200-6220
    MESGbmg1À60[email protected]
    offset 0d54f0-500
    MESGbmg100[email protected]
    now then this could directly tye into the reason for the crash computers hate infinity and they will try to work it out to the last number the same as the route of pie or 3.14594...... thus this could be what they are using to create the ap however i doubt that this is the only problem

    The bad Dump theory :
    Badlymcbald conclusion :
    <i>back up now then i noticed u lot have come back to one of the original hypothesis that this is a graphics glitch which is what it seems to be it is more then likely that we were originally thinking this yesterdayd as the crash occurs in 2 ways first simple black screen as the rom for what ever reason looses its ability to properly project its image every time this occur a dma 8bit error is show and generally the game can be reset and restarted the 2nd way the touch screen looses it capabilites and the next action causes errors to be found, all of the major address that have been looked are primarily used in graphics and screen projection i have gone through the code the only obvious thing wait a few days for a fresh rom to be dumped this one does appear to be glitched and is more then likely unfixable</i>

    =><i> There's no AP, it's a bad dump.</i>

    Why it cannot be a bad dump :
    - That would mean more than 4 bad dump : HG (U), SS (U), HG (F), SS (F) and another HG (F)
    - Why does the rom work on AKAIO ?

    But :
    - The AP maybe <b>cause</b> a bad dump ?

    Unreadable explanation of the bug by Badlymcbald, translated by MovieGuy :
    If you take the time to read the error that your reading as AP is the fact that there is an 16bit offset that is constantly being read as an 8bit.

    When this offset goes above the 8bit address its cause the black screen to appear.

    It is an error in the graphics engine of the game, every address that is looked at are addresses that deal in the graphical property of the game.

    04000600 is used to refresh the screen during entering of battle, 2260344 is primarily used to drawn sprites and set the writing.

    After the the other address that came up later server similar functions there is a constant message in desmume that there is an error in the lighting of the game that it is been shown as to bright.

    Now then if this wasnt a pure graphics error that it appears to be that it is AP the whole game would freeze including the sound.
    Thus it has to be that the AP was the original problem with loading up of the game originally, which was fixed.
    This was the original black screen, all the other freezes are occuring exspeciallly later on due to the nature of the graphics.
    The more objects and animation that is require to draw the harder it is for the game, to due this there that address is used more and more the address that is always seen in the crash.

    00056612 is used by the emulator itself not the rom.
    Note as well when the crash is seen in desmume you will also get crash reports from your computer indicating that various graphics files have crashed.

    Souldragon and me where looking at this yesterday but i pushed it aside originally trying to bypass thinking it was something that can be fixed.
    It cannot be fixed wait for a new version of the rom or continue in futility.

    Next possibility as to why this is occuring if it is not the roms its the emulators most have similar coding.
    Thats why it seems so random every1 assumes it is directly at 24 actions.
    Desmume experiences the crash at around 50 actions and nosgba crash at 60-75 actions, there is one thing ive yet to try that is to change the emulator on how it displayes light because this seems to direcly corralate the problem that the brightness of the game is directly what is causing the crash in desmume and nosgba

    Badlymcbald continue is explanation (Please somebody to translate) :
    i took a break and analyzed what is known so far now then the glitch occurs in the graphics error when the light is "turned off" by a cascading failure that is occuring at 56612 which is set as 07 nowthen leaving that alone there must be present a level that can be modified as the shader does automatically on screen reset of the sprites at offset 02260344 now then what is needed to be done is to locate the exact point at which the cascade begins as it begins as a dma error there must be a direct cause to this now then what is needed to be found is the exact point at which the dma is launched 1 of 2possibilities to the is that the ap is what is launching the dma event or that it is a general error in the reading of the rom by the various programs since not all programs the crash occurs at the same time the ap must not be the culprit as it whould have to have strict routine that is to be followed
    now ine direct line of this is the fact that 2 of the more used emulators have both and access of 50 graphics reset actions before failure while the average flash card system has less then that amount.
    if the failure can expediated by the gym badges as some claim then its should be exponential which it seems to be yet there are cases where individuals have made it to the elite 4 thus the case that badges as a possible way of locating the error seems flawed
    next the coding that streams itself seems to be in direct correalation to the errors thus by setting them to be skipped or over ridden caused complete failure of the rom as they are needed by the graphics engine to run the game so they cannot be directly changed.
    hypothesis-
    game mechanics operate a higher level due to the direct refrence of screen changing the emulators use more ram to calculate the exact pixel ratio needed to run the next screen it is in there that the problem lays if we can directly influence the outcome of those calculations being made it should be able to directly prevent the freezing from occuring any ideas on how to due this

    => Sure it's a graphic engine bug
    The RAM overflow theory :
    someonewhodied :
    <i>im pretty sure that its because of a ram overload...Normaly, no$gba takes 146,284 k of ram (on my PC) but every time i enter/exit a building, the ram spikes up to 297,186 k and then drops back down...maybe thats whats causing the freeze on flashcarts??? on the jp version, the ram only spikes up to 211,912 k.</i>

    Well, that makes a lot of Spoiler.



    From Ribesg.

    People didn't get the "no opening more threads" message so I will warn if more are opened. There is a reason to why threads get closed, please accept it until a mod deems its worthwhile to continue.
     
    colt05 likes this.
  2. rockstar99

    Member

    Joined:
    Dec 3, 2008
    Messages:
    7,372
    Country:
    Canada
    no u
    go to the other thread
     
  3. eugene222

    eugene222 GBAtemp Regular
    Member

    Joined:
    Sep 12, 2009
    Messages:
    183
    Country:
    Gambia, The
  4. Kroatien99

    OP Kroatien99 GBAtemp Regular
    Member

    Joined:
    Mar 14, 2010
    Messages:
    140
    Country:
    Croatia
    It is closed.
     
  5. rockstar99

    Member

    Joined:
    Dec 3, 2008
    Messages:
    7,372
    Country:
    Canada
    lol it ran outta pages [​IMG]
     
  6. Bumpman

    Bumpman GBAtemp Regular
    Member

    Joined:
    Jan 21, 2010
    Messages:
    252
    Country:
    United States
    As soon as you start the game, search for the value 18 in a debugger, to see if it decreases as you do events
    18 = 14 in hex
     
  7. atenzor

    atenzor Advanced Member
    Newcomer

    Joined:
    Mar 14, 2010
    Messages:
    78
    Country:
    Canada
    Did anyone tried this (cause this is strange):

    The US roms have two EXTRA files added:
    \heartgold\data\a\2\6 - - 4
    \heartgold\overlay\overlay_0128.bin

    What happens if you remove them and re-pack the ROM? Does it fix the problems?
    Can someone try or did someone try?
     
  8. Donixs

    Donixs GBAtemp Regular
    Member

    Joined:
    Aug 12, 2009
    Messages:
    132
    Country:
    United States
    Already been tried, it failed.
     
  9. Kroatien99

    OP Kroatien99 GBAtemp Regular
    Member

    Joined:
    Mar 14, 2010
    Messages:
    140
    Country:
    Croatia
    You try it.
     
  10. Shady Guy Jose

    Shady Guy Jose GBAtemp Fan
    Member

    Joined:
    Sep 14, 2008
    Messages:
    497
    Country:
    Portugal
    Maybe try that and THEN apply Japanese AP patches?
     
  11. Kroatien99

    OP Kroatien99 GBAtemp Regular
    Member

    Joined:
    Mar 14, 2010
    Messages:
    140
    Country:
    Croatia
    Has anybody compared Jap AP and US AP.
    Maybe thay have used different AP on same place.
     
  12. atenzor

    atenzor Advanced Member
    Newcomer

    Joined:
    Mar 14, 2010
    Messages:
    78
    Country:
    Canada
    I discovered something else, the ARM9 and some other binary files have different Byte size than JAP roms?
     
  13. Bumpman

    Bumpman GBAtemp Regular
    Member

    Joined:
    Jan 21, 2010
    Messages:
    252
    Country:
    United States
    Yes, many times
    The JAP AP is extreamly sporadic, while the US AP happens after 24 events befor eyou get the 1st badge
     
  14. Shady Guy Jose

    Shady Guy Jose GBAtemp Fan
    Member

    Joined:
    Sep 14, 2008
    Messages:
    497
    Country:
    Portugal
    Yes. They're different. US one is the same as KOR one, and that one hasn't been cracked yet.
     
  15. atenzor

    atenzor Advanced Member
    Newcomer

    Joined:
    Mar 14, 2010
    Messages:
    78
    Country:
    Canada
    Yes, maybe...
     
  16. eugene222

    eugene222 GBAtemp Regular
    Member

    Joined:
    Sep 12, 2009
    Messages:
    183
    Country:
    Gambia, The
    Kroation pls update the first post with this:


    - It works 100% on AKAIO RC 1.6
    - Tried to apply the JP fix : Fail
    - Tried to move US files to JP one, to use that JP fix : Epic Fail
    - Tried to patch ARM7 with JP HG/SS one, Platinum one, D/P one : Fail
    - Tried to patch ARM9 with JP HG/SS one, Platinum one, D/P one : Fail
    - Defined "events" as "when you enter/exit a building/stairs/door or save or encounter a wild pokemon or start a trainer fight or end a battle (get back to the game) or when you end a cutscene"
    - The number of events "needed" before the game freezes depend of the number of badges you have
    - Replacing all the 00 00 00 46 in the rom by 00 00 00 00 : Fail
    - Tried a lot of AR codes : Some from JP, K fix... No one working
    - Unchecking DMA and reset : Fail
    - Inserting patched JP overlays : Fail, but increase the number of events before freeze
    - Deleting overlays : Fail
    - This match with the equation -(1/2)*x^3 - (5/2)*x + 24
    - You replace x with the number of badges you have, and you find : 0 badges 24 events, 1 badges 21 events, 2 badges 15 event, 3 badges 3 events, more badges you have a negative value and the game do not want to load the save
    - With this "events" needed we can think there is a countdown, each event add -1 to a var who doesn't support value below 1, that can explain the countdown crashes and the fact that the game won't start if the value is negative. If it's right, a AR code who lock this var must work...

    This list was made by Ribesg
     
  17. banditkeith

    banditkeith Advanced Member
    Newcomer

    Joined:
    Sep 12, 2009
    Messages:
    67
    Country:
    So is this the discussion thread until that one unlocks?
     
  18. Kroatien99

    OP Kroatien99 GBAtemp Regular
    Member

    Joined:
    Mar 14, 2010
    Messages:
    140
    Country:
    Croatia
    That will maybe work.
    Has anybody tried it.
     
  19. Bumpman

    Bumpman GBAtemp Regular
    Member

    Joined:
    Jan 21, 2010
    Messages:
    252
    Country:
    United States
    Yup, and if mods can merge, they should merge this one with that one afterwards
     
  20. Luna

    Luna GBAtemp Fan
    Member

    Joined:
    Mar 18, 2009
    Messages:
    346
    Country:
    Has anyone seen this yet?

    EDIT: Added the AR code.
     
Loading...

Hide similar threads Similar threads with keywords - SoulSilver, HeartGold, problems

Thread Status:
Not open for further replies.