I took a gamble on some fishy-looking proto boards from Japan. The boards themselves have no markings, and the stickers look brand new, but all of the EPROMs were manufactured back when these games were being worked on, so I figured what the heck. Clockwise from the top-left we have Stargate, Joust, Soccer, and Hyper Olympic.

Anyway I dumped them and, hey, all but Soccer are different than the final versions. Which in itself is REALLY cool, but I want to call your attention to Joust.


Joust (along with Stargate and Millipede) were ported from the arcade originals way back in 1983 - the launch year of the Famicom - by HAL Laboratories, as contract work for Nintendo. They were, at the time, first-party games. They were made as part of Nintendo's pitch to Atari in America: if you're unaware, before the "Nintendo Entertainment System" was even a concept here, Nintendo was trying to get Atari to distribute the Famicom as its follow-up to the Atari 5200 (the 7800 was in development at the time, but wasn't announced).

Anyway, that deal fell through, and all three games were shelved. Fast forward all the way to 1987, and HAL eventually self-published its port of Joust, so technically this game was released. However, based on the other prototypes this seller had (the majority of them first-party games from 1983-1985), these appear to be from the time that it was a first-party Nintendo title. So, you might call this a prototype of an unreleased first-party Nintendo game, which is pretty cool.

What's even cooler is that it is early code! There are clear differences, including some English typos that were fixed in the final game. So, one might argue, it's an incomplete prototype of an unreleased, first-party Famicom game from its launch year. Which is amazing!

But here's the kicker: Joust was programmed by Satoru Iwata, who eventually became the president of Nintendo. It was, in fact, the first game he ever coded for Nintendo. Which means that in addition to all of the above, this is also the earliest documented Nintendo code by the friggin' legend that eventually oversaw the entire company.

Anyway. I tweeted a bit about this yesterday, and Polygon ended up writing a story, so you can see more there: http://www.polygon.com/2016/8/16/...

Wow. That's kinda amazingly cool! A very neat piece of history.

I blinked in disbelief until it all sank in. Amazing!

It's amazing at how much undocumented treasure is still out there. Thanks for sharing this here, Frank, and being so committed to game preservation.

Neato bandito!  

that's awesome!

Incredibly cool! Thanks for sharing.

I had no idea about Nintendo's history with Atari. Fascinating!

Was the Famicon deal supposed to be just a North America thing or what?

Wow, I went from mildly interested to "Shut Up and Take My Money!" by the time I was getting to the 3rd paragraph. Nice find.

Amazing find. Good job!

How do you find this stuff?

This is awesome! Iwata will be forever missed.

Very cool, and thank you for providing the background. What a sentimental treasure!

What an amazing piece. Thanks for the backstory  

Am I reading it wrong, or were the Joust EPROMs made in the middle of 1984?

Never seen Famicom prototype PCBs like that, look more like pirate PCBs. Strange.

The "15" over one of the middle pins actually reminds me of an N64 game pak PCB I did trace repair on yesterday. It just smacks me as being a bit "new," but the OP himself acknowledged that appearance.

OP: doing a comparison between final ROMs, are they simple text edits only or does any of the actual game code appear to be tweaked?
 

Yep, they're from 1984. My guess is that the EPROMs themselves are actual vintage, and the boards & stickers are newer (the stickers are pretty much flawless).

The same seller sold 16 of these, here's what I can suss from the auctions:

- All of the EPROMs were manufatured between 1983 and 1985
- 15 of the 16 boards were exactly like mine
- The remaining board had Nintendo 1983 branding on it, and a part number that isn't documented anywhere online

We don't know enough about Joust's development to know when it started and ended, but having examined it, I'm confident the code on these EPROMs is legitimate. Maybe the source of this stuff had the one real board that they clipped EPROMs into (they're all NROM games), and when it came time to sell, they put the ROMs on their own boards? No idea.

Lots of code tweaks, but nothing that obviously jumps out. The tiles are identical. It's not super early or anything, it's the complete game, I'm guessing all the differences are polish/balance/bug fixing.

Man, Frank, awesome work! You're a wealth of knowledge. Finish that book you occasionally mention, so that I can quit re-reading the old LL articles every year or two  .

What does the title screen look like?

Edit:  Doesn't seem as though that would tell us much.

Didn't the possible Nintendo/Atari deal occur earlier than the 20th week of 1984?  Why would the chips have been burned so late for a potential 83/84 deal?

The title screen is identical to retail, which says 1983 Atari on it. The code is absolutely earlier overall, but like I said, not by a lot.

I don't know the entire history of this product or the Atari/Nintendo deal, only what's online and in Game Over, which doesn't give reliable dates. We know from an existing Atari memo that it was still being discussed as of mid-June, 1983, but that is the only hard date we have. If you've got insight that we don't, please share. All I can do with the facts that I have is play the imagination game and give you a list of some scenarios I can dream up. Off the top of my head:

- The deal was still being discussed with Atari through that period in 1984.

- After the Atari thing, Nintendo was demonstrating games to some other distributor for a deal that isn't documented or generally known.

- After the deal to distribute the Famicom in the U.S. fell through, Nintendo was still trying to publish these games in Japan under a license from Atari.

- A Nintendo employee was burning ROMs for themselves to take home and play for fun.

- A HAL employee burned data of an unreleased game to demonstrate their work to a client.

- There was a piracy scene among developers who traded ROMs to each other.

- HAL was evaluating its past work, in the lead up to their self-publishing the game in 1987.

- Yamauchi came into the office drunk and demanded his subordinates dig out their unreleased games and make him copies or he'd have their families killed. He mentioned something about selling them on "eBay" because he was secretly a time-travelling wizard, but nobody knew what that word meant.

- This whole thing is an incredibly elaborate scam from the seller. They got every meticulous detail down, including breaking the games just slightly, finding vintage EPROMs, and even leaving Soccer as the final version just to add a little credibility to the whole thing, but they missed that one thing.

- Despite the 1983 copyright, Iwata developed the demo in 1984.

The PCBs, stickers on the EPROMs, and manufacture dates on the EPROMs are all odd. That's fairly indisputable. The code gives the carts some credibility, so that's worth noting as well.

For someone so adamant about historical accuracy, I would assume you'd look into here closer. I'd be curious what the seller says about they obtained the games, the origin of the stickers, etc.

I've had my fair share of protos come through my hands, there are some things that are quite obviously strange about these carts. It's worthy noting and looking into.

Burning games for AVS demo? Does that line up with any CES?

The final game didn't come out until October 30, 1987, so I don't find 1984 EPROMs odd. A million things could have happened between 1983 and 1987 to make these EPROMs exist, as I've done my best to illustrate to you. I'll give you another scenario: this is an interim build between the original "pitch version" and the one that was released, maybe it should be classified as a prototype of the HAL version instead of an internal Nintendo prototypes. Feel free to do so if you'd like to.

As for the boards and the stickers, I agree with you, and said so. I have also "had my fair share of protos come through my hands," for almost 18 years now, which you are aware of. So I'd like to think I know at least as much as you do at this point. And my opinion, as a person experienced in these matters, and with the limited facts on hand that we have, is the same as what I said it was in the first post in this thread: these are most likely real vintage EPROMs with real vintage code that were soldered onto newer boards with newer stickers, for reasons unknown. And until someone can point me to some other source for early ROMs of these games, that is where my opinion is likely to stay. Whatever you want to classify the items as, "prototypes" or bootlegs or loose EPROMs or whatever, I don't really care: the code appears to me to be authentic, and the rest of it is a mystery. If that's all we ever know about these items, I'm more than okay with that. I'd love to know more, and I'm doing my best, but that may never happen.

Sure, there's another possibility. AVS was shown January of 1985, so it fits. But like I was trying to illustrate, we can make up scenarios all night long, it's not going to solve anything!

Unless you meant your AVS, in which case, I would advise you not to demonstrate your product at CES, as the exhibitor fees are enormous.

Sorry for the triple post, but I want to give Mark some credit here, now that I've thought about it some more: it was wrong of me to actually call this a prototype from the time of the Atari pitch, there is NO WAY I could substantiate that. Obviously I'm excited and want that to be the case, and may have been clouded by that. It still could be, of course, but it's wrong to say that specifically, like I did in the title here.