Bounty-driven project to reverse-engineer Nintendo DS Wi-fi hits first milestone; blogger complains

by Jason McIntosh

Related link:

I was getting ready to post something mildly critical about StoneCypher's Nintendo DS Wi-fi bounty when I hit the page just now and noticed that one Stephen Stair seems to have claimed the first chunk of the bounty by posting a reference to the DS's Wi-fi hardware. (The remaining slices to go those who can use that information to implement things like a TCP stack.) I am no low-level network hacker and balk at these long tables of hex-addressed I/O maps done up in 10-point Courier, but it certainly looks clean and impressive, at least.

Bit of background: the Nintendo DS is one of the current generation of portable video game systems, and the one that, due to its relatively low price and high innovation, did in fact leap into my arms the last time I walked near an EB Games shop. It boasts built-in 802.11b support, but Nintendo doesn't make documentation available for it (or any other aspect of the DS) to anyone except for dues-paying licensed developers, leaving the home-brew hacker community in the cold. And you know how they get, when denied access to such a tasty technological morsel like the DS. So earlier this year, that StoneCypher fellow organized that bounty project aimed at reverse-engineering the little system's WiFi hardware enough so that hackers could start making their own networked and even Interner-capable DS games.

This is a noble cause, and I am well pleased to see that it's apparently hit its first major milestone. However, I do not wish to spare this project from the wrath light drizzle of my mild criticism, which I believe remains valid despite this achievement. In particular, I wonder how much duplication of work happened with people working in blind parallel with Mr. Stair. From all that I can tell, Stair worked alone, but I doubt he was the only party vying for the first slice of the prize.

I am not really sure one way or the other, though, and that actually points to the problem I see: because this effort was framed as a competition rather than an open, cooperative project, every developer (or discrete group of developers) who participated had an incentive to not share their work, or even communicate or assist each other in any way. I wasn't part of the bounty-hunt so I can only speculate how it went down, but I am guessing that the main page's lack of links to anyone's individual efforts before now (and continued lack of links to anyone building upon Stair's map to claim the remaining bounty) are indicative of the general lights-out nature of the whole deal.

How much faster would this milestone have been reached if someone had organized the Wi-Fi hacking as an open (and open-source) project instead? It's true that nobody would likely get an immediate cash payout from it, but I bet that this wouldn't stop anyone. The money, after all, is not the reason to accept this sort of challenge. $1,400 seems like an impressive amount, but with the amount of work he must have put into reverse-engineering all those register locations, I bet the recipient could have made more money performing rote data-entry or even burger-flipping over the same amount of time.

I cannot be too wrathful because I know damn well that if this bounty-based effort reaches its ultimate goal I'll celebrate alongside every other game-hacker hobbyist. But even then I'll still wonder if it was an effort that was held back by months due to a failure to apply open-source programming principles to a very cool and edgy project that could, I'll wager, really benefit from it.

Eh, nuts to this. I'm off to play more Meteos.


2005-11-02 00:00:07
solo vs cooperative
While an argument can be made for cooperation, I wonder how much of that potential gain in project efficiency would be lost on coordination (and yes competition) among members.
2005-11-02 04:25:18
Maybe yes, maybe no...
As Malcolm Gladwell evidenced in "The Bakeoff"[1] open source groups sometimes don't perform as well as traditionally formed groups. Gladwell suggests that this is because of the friction between the expert members of the group, something that might have occurred with the open source hack you suggest. Also, Gladwell suggests that open source projects are good at "chasing taillights" and not so much at innovation. (But arguably, the DS hacking is chasing taillight.) Also, if you take some of the ideas from James Surowiecki's "Wisdom of the Crowds" [2] benefits come not from one large group but lots of little groups working on the same problem and producing many different solutions. And of course, never discount self-interest. Even thought the cash prize might pale in comparison to the work, if the work is more like a hobby to you, it's like getting paid to have fun.