iceissocold said:
chartube12 said:
how do we know the pokewalker uses ir signle?
Doesn't the ds also have ir built into it's wifi card? Witch is what is used for direct play and the ds chat thing?
Just look at Atashi's post
http://gbatemp.net/index.php?showtopic=215641 you can see the IR sensor component in both pictures, located near the top of the retail DS Gamecard.
Something like this might be possible but would require the retail gamecard and a pokewalker. Trying to get an IR sensor from a PC to talk to the one on the DS cart would make things easier IMO.
I looked at the pix and read the description. This just SCREAMS memory mapping of the IO to me; I did assembly language programming on the Apple 2 way the hell back in the day and its CPU had ZERO I/O ports - 100% of the machine's I/O was accomplished by using memory addresses reassigned to devices.. read from $C030 and the speaker clicks, reading from another series of addresses would turn the disk drive on or off, writing to a third outputted a byte, etc ...
MMIO also naturally lends itself to being used anywhere memory might be found. Like, in a cartridge. Why waste extra pins for ports, or a cart Nintendo has already gone out of its way to make smaller than previous carts... additionally, the fact that none of the reader/writers yet work with HG/SS lends credibility to this observation (PMIO wouldn't get in the way of accessing memory, but MMIO left switched to the wrong state can make memory inaccessible.)
The fact that a 1MB file full of nulls is located by the other hardware is also valuable: it indicates that when in some other mode, there seems to be a 1MB space. Is it 100% nulls, or are there a few values that aren't null? If so, focus your attention on those locations and how interacting with a physical pokewalker changes them (does d/ling the 1MB of 'dead space' always produce the exact same file?)
---
From watching the thing blink as it transfers (via camera), I've come to a couple
GUESSES:
Pokemon aren't stored on the device. Not because they can't be, but because it would be pointless anyways, the device doesn't have the game mechanics to deal with it; and the primary pokemon can't be "traded out" - in fact, Nintendo even went out of its way to ensure that the cart could be reset if the pokewalker was lost - doing so would explicitly make cloning possible unless the pokemon wasn't on the walker in the first place. They seem to be aware of these sort of issues, so ...
The device doesn't have a pokedex.
The paths are relatively simple, and consist of less data than B&W images of your pokemon walking..
The device doesn't store your pokemon - but it does download several really nice bitmap images of it. Probably RLE graphics.
The device doesn't let you capture pokemon, just names & ID values that are sent back to the game.
...So, you can't catch Flychu in the Refreshing Field with a hack because there won't be any way to communicate the idea in the first place; any pokemon you catch doesn't get stats, moves, etc until it returns to the DS. So to hack a Flychu with infrared, you will at the very least need to have the ability to return from the Yellow Forest / i.e. you still need the event.
... All of this is resent at each download.
I would also guess that the processor inside the unit uses numeric registers that contain a single bit, rather than the usual 8,16,32, or 64 seen in desktop systems.
I make these guesses because it would be both easier and cheaper to manufacture the device that way, than in a way that involved it manipulating the actual data of the pokemon itself, and the predicted behaviors that would result match how the thing works in real life very well.
