On Wed, Jul 27, 2011 at 5:38 AM, Steve Gulick <[email protected]>wrote:
> Hi Eric, > > Yes, the way you diagramed it is what I mean. > Okay. It is as I suspected. You really don't want to do this. It has numerous problems. You are replacing a physical media by a radio media that has packet size issues, flow control issues, and buffering issues. Originally the mote that I'm working on had a radio module called the Wi.232 which was intended to do exactly what you are talking about. After analysing the characteristics of the design we abandoned it. Too many problems, not enough payback. I'd recommend you do the same and find another way to solve your problem. One option is you make the mote run as a serial/radio ipv6 gateway and run PPPd on the Linux box. Then the radio medium looks like another IPv6 network. While this too has issues, they are very different than the issues I mentioned above. The network stack takes care of at least the framentation problem. Not sure about flow control (TCP6 handles flow control, but I don't think that UDP6 does, so is still a problem there). But bottom line is pretending that two motes with radios is the same as a serial line is really a bad idea. Packetization, buffering, etc destroy the physical medium assumption. I'm off back to what I should be doing. YMMV eric > > What I wanted was a mote connected to a PC that would look to any > Linux application (e.g. minicom) like a serial port and would be > connected to over a radio link to a similar mote and PC appearing as a > serial port to any arbitrary Linux application. > > Like the serial line replacement mode available on the XBee radios > from Digi. There the packetizing and retries are handled transparently > on the XBee module. > > Or another example would be the Lantronix Matchport serial to Wifi module. > > Steve > > > > > On 7/27/11, Eric Decker <[email protected]> wrote: > > Huh? > > > > First you talk about a serial line replacement and then you talk about > about > > using the serial forwarder. > > > > I'm confused as to just what exactly you are talking about. > > > > So I think what you are talking about is something like.... > > > > PC ---- serial --- mote ---- radio --- mote --- serial --- PC. > > > > > > Is that what you are talking about? > > > > Sure there is a way to do this, you figure out how you want them to > talk > > and then implement it. > > > > Look at how motes can talk via radio, what the characteristics of that > > communications path is, ie. packet size, packet format, flow control etc. > > > > And then you figure the same thing out for the PC to mote across the > serial > > link. And what that means for what the program needs to do running on > the > > mote. > > > > But why are you doing this? A bettter place to start is what problem > are > > you trying to solve and is it worth the effort to implement the way I've > > described. > > > > Frankly it seems odd to me. why bother? > > > > eric > > i've done networking for years. > > > > On Tue, Jul 26, 2011 at 5:24 PM, steve <[email protected]> > wrote: > > > >> Hello, > >> > >> I would like to create a serial line replacement between 2 PCs using 2 > >> motes. > >> Can anyone suggest a way? > >> I was think of something like a serial forwarder attached to each of the > >> PCs > >> > >> Thanks > >> Steve > >> > >> _______________________________________________ > >> Tinyos-help mailing list > >> [email protected] > >> > https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help > >> > > > > > > > > -- > > Eric B. Decker > > Senior (over 50 :-) Researcher > > > -- Eric B. Decker Senior (over 50 :-) Researcher
_______________________________________________ Tinyos-help mailing list [email protected] https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
