View previous topic :: View next topic |
Author |
Message |
theclaw
Joined: 19 Dec 2010 Posts: 17
|
Posted: Tue Mar 06, 2012 1:55 am Post subject: Were WWF No Mercy 1.1 (NTSC) carts actually issued? |
|
|
As is common knowledge, many copies of WWF No Mercy for N64 have a fatal recurring save deletion glitch. All save data stored on the cartridge forever vanishing in an unpredictable instant. Gone. End of story.
Not something any serious gamer who's played the game extensively would just forget. Years experiencing repeated total loss of unlockables, created wrestlers, etc.
So I'm wondering if the rumored fixed NTSC cartridges were in fact created. I haven't found many leads except this article from 2000.
http://www.gamespot.com/news/wwf-no-mercy-bug-riddled-2656988
PAL is no question, we already know those exist. The PAL 1.1 rom was dumped and there's even photographic evidence of PAL carts with a revised front label. (though I haven't heard whether PAL 1.1 fixes the save glitch!) |
|
Back to top |
|
|
Nosuch
Joined: 01 Dec 2006 Posts: 107
|
Posted: Tue Mar 06, 2012 10:29 pm Post subject: |
|
|
My copy is the revised 1.1 version so I can vouch for their existence. I contacted THQ way back when and they exchanged it for a fixed version. |
|
Back to top |
|
|
theclaw
Joined: 19 Dec 2010 Posts: 17
|
Posted: Tue Mar 06, 2012 10:52 pm Post subject: |
|
|
In that case perhaps dumping the cart's rom may tell us more? No one's conclusively verified the technical reason this glitch happens. |
|
Back to top |
|
|
spoondiddly
Joined: 09 Oct 2005 Posts: 10
|
Posted: Wed Apr 04, 2012 2:48 pm Post subject: |
|
|
Actually, if I could get ahold of both versions I could work out where the glitch lies. Might even be patchable with codes.
Strangely many 1.1 NTSC versions were never dumped. It wasn't until a couple years ago that the v1.1 Ogre64 saw light, and that had, besides some silly glitches and japanese, a persistant inventory incrementation bug that would (without user management) eventually crash the game. Totally exploitable too, concidering you could allow it to increment through the item list to gain rare/expensive/unique stuff.
For giggles, here's No Mercy's decompression routine in psuedo-pythonese:
Code: | # filesize is 3-bytes, skipping first which is compression indicator
# 0x0 1 00 if compressed
# 0x1 3 filesize, big-endian
size = int.from_bytes(data[1:4], byteorder='big')
d = _grab(data[4:])
while len(output)<dec_s:
cmd>>=1
if cmd<2:
cmd = next(d)|0x100
if cmd&1:
output.append(next(d))
else:
p = next(d)
l = next(d)
p|= ((l<<4)&0xF00)
l&=0xF
p+=0x12
p&=0xFFF
# funky correction here
c= len(output)&0xFFF
p-=c
if p>=0:
p-=0x1000
p+=len(output)
# loopus
for i in range(l+3):
v = 0 if p<0 else output[p]
output.append(v)
p+=1
return output |
Filelist clocks in just short of 19500 entries.
Strangely, a lot of hackers patch against v1.0s even when later ones are available. It doesn't make a whole lot of sense honestly. Worst case was the run of people patching against Perfect Dark v1.0 with the murderous save file bug. Aegh! |
|
Back to top |
|
|
theclaw
Joined: 19 Dec 2010 Posts: 17
|
Posted: Thu Apr 05, 2012 1:38 am Post subject: |
|
|
That's good to hear. Photos of 1.1 would go a long way. Especially the PCB for any internal difference. |
|
Back to top |
|
|
spoondiddly
Joined: 09 Oct 2005 Posts: 10
|
Posted: Thu Apr 05, 2012 9:11 pm Post subject: |
|
|
There wouldn't be an internal difference. N64 carts are very standardized.
Comparing ROM images (aka dumps) would reveal whatever fixes occured. A patch could also be made by examining the fix in a different region and porting it, but a direct correspondance is much easier to deal with instead of trying to account for differently mapped addressing and whatnot.
There were a couple titles from that span of time with murderous save issues. Must have been a problem with a subversion of the standard libraries. |
|
Back to top |
|
|
theclaw
Joined: 19 Dec 2010 Posts: 17
|
Posted: Fri Apr 06, 2012 10:13 am Post subject: |
|
|
I meant things like serial numbers on the ROM chips that may stand out as clear 1.1 identifiers.
Still don't get ahead of yourself spoondiddly. I haven't yet been provided with entire proof it exists. User reports and comments yes, including in this topic. But nothing that jumps out proudly shouting and waving flags "I am a WWF No Mercy 1.1 NTSC cartridge!!." |
|
Back to top |
|
|
spoondiddly
Joined: 09 Oct 2005 Posts: 10
|
Posted: Fri Apr 06, 2012 6:42 pm Post subject: |
|
|
It probably wouldn't be stamped on the chip. There's two ways to tell a version difference:
- On the cart label it will list the ID as such: NUS-NW4E-USA-01
- In the ROM header at 0x3B it would read: 4E 57 34 45 01 (NW4E\x1)
Not all will follow the cart label standard (very few exceptions though) but all internal headers will. |
|
Back to top |
|
|
theclaw
Joined: 19 Dec 2010 Posts: 17
|
Posted: Fri Apr 06, 2012 7:30 pm Post subject: |
|
|
Yeah we might be getting somewhere. I've just now dug up info on a potential ROM revision!
http://datomatic.no-intro.org/index.php?page=show_record&n=0862&s=24
http://datomatic.no-intro.org/index.php?page=show_record&s=24&n=0882
CRC differences don't tell me much, since I'm not technically inclined. It's been named "Rev A" by the dumping scene and can be found on ROM sites.
I'm not sure whether this one is legit or fixes the save glitch. Or in the event it's legit, reached cart format. Legit could mean leaked proto/beta from another source.
Possibly even a third version that's in addition to what Nosuch mentioned? I wouldn't rule out the full story here ending up weirder than fiction. |
|
Back to top |
|
|
spoondiddly
Joined: 09 Oct 2005 Posts: 10
|
Posted: Sat Apr 07, 2012 1:22 pm Post subject: |
|
|
If they follow convention Rev A would be v1.1. Prototypes are always clearly marked proto and demo carts the same. It only gets really complicated with Turok II, with multiple prototype and demo versions of questionable age. Seriously, proto and demo Turok II carts are probably more common than retail.
On a cursory glance memory has been shifted around a bit. It's mostly in the setup right after the bootstrap. I'll really take a good poke at it though and see what's up.
On a slightly less cursory glance there are similiar bug fixes in there as between PAL. A few function differences and removing reliances on tables. That frees up quite a bit of memory and it looks like they used it for a larger mempak buffer (among other things). |
|
Back to top |
|
|
theclaw
Joined: 19 Dec 2010 Posts: 17
|
Posted: Mon Jul 23, 2012 7:43 am Post subject: |
|
|
Any new info found? Whatever you guys can tell me about identifying it would be welcome. That's going to be key to setting in stone the answer of this topic's question.
Rather not have to pick up stacks of the game, then play each one long enough to see whether saved data vanishes. I'd go buy a PAL N64 and its easy to tell apart PAL 1.1 before wasting resources that way. |
|
Back to top |
|
|
Skrybe Staff
Joined: 28 Aug 2003 Posts: 1610
|
Posted: Mon Jul 23, 2012 5:35 pm Post subject: |
|
|
I haven't looked at the files personally, but it's very likely the Rev A version you linked to is the fixed copy. If you have any doubts, someone at the No-Intro forums might be able to tell you where that dump came from. |
|
Back to top |
|
|
theclaw
Joined: 19 Dec 2010 Posts: 17
|
Posted: Mon Jul 23, 2012 5:47 pm Post subject: |
|
|
Sounds good. At this point if I did identify and track down a fixed cart, it might still lose data. We haven't actually verified the primary bug was squashed. |
|
Back to top |
|
|
Nosuch
Joined: 01 Dec 2006 Posts: 107
|
Posted: Mon Jul 23, 2012 11:56 pm Post subject: |
|
|
I dunno, I'm pretty confident that they did squash it. Like I said above, I've had a fixed copy since back when the game was current and the bug hasn't reared its ugly head even once. |
|
Back to top |
|
|
theclaw
Joined: 19 Dec 2010 Posts: 17
|
Posted: Tue Jul 24, 2012 12:23 am Post subject: |
|
|
Nosuch wrote: | I dunno, I'm pretty confident that they did squash it. Like I said above, I've had a fixed copy since back when the game was current and the bug hasn't reared its ugly head even once. |
Well anything at all you can tell me different looking from the original, will be the best clue we have at this point. There might be other overlooked details like a number stamped into the back label.
A revision should have some indicator the manufacturer used to separate them. Otherwise there's greater chance they sent bugged carts back to people who exchanged. |
|
Back to top |
|
|
Nosuch
Joined: 01 Dec 2006 Posts: 107
|
Posted: Tue Jul 24, 2012 5:34 pm Post subject: |
|
|
I don't have a way to compare it against a bugged copy but, at a glance, the only thing I can really tell that would for sure be different is the product code having a -1 at the end. Still, I scanned them anyway. (The label will show some discoloration from years of use; don't mind that.)
Front
Back |
|
Back to top |
|
|
theclaw
Joined: 19 Dec 2010 Posts: 17
|
Posted: Tue Jul 24, 2012 7:41 pm Post subject: |
|
|
Thank you. Knowing about -1 changes the whole story. From a question of existence, to no longer unique, within hours.
I found a highly probable 1.1 on ebay thanks to your update. (I'd first try to find one better condition than this written on sample)
http://www.ebay.com/itm/271022318393?_trksid=p5197.c0.m619 |
|
Back to top |
|
|
|