> > > So IrDA SIR mode uses async mode which and has to be framed and stuffed in
> > > software. MIR and FIR mode are however handled in hardware, and those chips
> > > has very frame oriented interfaces. So we don't really need any
> >
> > >From what I seen (pc87108, tir2000), even at MIR/FIR speeds hardware
> > looks like serial with length-of-packet added. I actually implemented
> > tir2000 FIR driver as tty - it adds frame begining/end when it
> > absolutely musts. Being tty has certain advantages in my eyes: most
> > interesting is "i'm able to test it with simple commands like cat and
> > echo". Also, being tty I could try to run something else than irda
> > over that line (for example scarab).
>
> Being network device also has certain advantages since you can get direct
> access to the driver through the packet interface. Implementing the
> irdadump program (works like tcpdump) was very easy since I got almost
> everything for free.
You are probably right. What I want is generic tty-over-netdevice
layer. And it may not be such a good idea for multipoint networks.
OTOH, radiomodems _are_ multipoint still they look like tty => you can
put drivers into userspace.
> I guess that the issue of being char or net device is really about being
> point-to-point or more like a multiple access MAC layer. IrDA is both and
> can function in connected reliable, unconnected reliable, connected
> unreliable, and unconnected unreliable mode. IrDA v2 known as AIR will also
> support multipoint which is not that easy to put over the TTY
> abstraction.
As long as multipoint is done in software (does irda do it this way)
like in radiomodems, and not in hardware like ethernet cards, putting
IrDAv2 over TTY will be ok.
> Some people complain about the many and complicated layers of IrDA, so I
> really don't understand why the same people wants to have the TTY subsystem
> underneath ;-)
1:0.
It is simple. It is so that I could take work of people doing
radiomodems (scarab), and use it with FIR devices to gain multipoint
4MBPS connections ;-). It's true that going to userland at speeds
~4MBPS may not be such a good idea... And it is so that I could try to
develop my own IP-over-infrared encapsulation, and do it completely in
userspace, and _still_ have it working with 4Mbit FIR devices.
Pavel
-- The best software in life is free (not shareware)! Pavel GCM d? s-: !g p?:+ au- a--@ w+ v- C++@ UL+++ L++ N++ E++ W--- M- Y- R+- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/