rconty wrote:I'm interested in.
Of course Romuald, you are more then welcome! 
The Mega board seams very nice indeed, though since I'm not really a hardware guy I hope we can group up and design the board all-together. The LCD can wait, it would be nice for "future" testing, but most really interesting standalone embedded devices have at least a small display. Besides of course, that it is sw33t for debugging
rconty wrote:But before choose the hardware board, could you give some examples of use case. It will help to determine many things like required pins number, MCU speed and connectivity.
I would say, it should be able to run at a reasonable embedded speed to do some crypto (AES, RSA, Elliptic Curve). Or just communicate to a SAM (Secure Access Module) in form of a smartcard, which could do those kind of operations as a secure co-processor. But there could be other requirements involved. I would say we try to sum them up to make a priority list (ala MoSCoW)
rconty wrote:You can also think about supporting I2C directly into libnfc. Many embedded devices run on linux and have I2C bus (by hardware or using linux's integrated bitbang driver), so for example you could revive some old routers and made NFC-capable hardware.
If we have the hardware, don't worry about the software, that is the place where my magic starts 8)
No serious, I guess we can bring this to a reasonable level. If this includes an I2C stack, internal low-level driver, I see no obstacles at all. The (linux)routers could be useful, but they are pretty big, and since I see a widespread and new market here, there probably will be some nice investment available.
It is about time that someone comes up with a reasonable open-source device. I have seen now so many insecure access-control, public-transport, library, parking-lots which all depend on close-sourced obscure methods, let's start the real NFC revolution and bring finally something worthy to the daylight.