Re: LZ: Teleconferece
[Prev][Next][Index][Thread]
Rob Taylor wrote:
>
> > > > Now, we believe that if several TIs are set on the same freq that they
> > > > should be able to comunicate together at once.
> > > >
> > > > .. snip ..
> > > >
> > > > Now I am not sure if this will work or not since (to my knowledge no
> > > > RT-LiNKS have been made by anyone but the author) no one has made 4 to work
> > > > together.
> > >
> > > This should not work. The TI does not use a signal that would allow for
> > > this. It is instead a simple on off type of signal designed for two
> > > calculator use. Simply putting more on the same freqs would not be any
> > > more successful than putting splitters on the regular link cable.
> >
> > Well, with the standard system routines, this would be correct, but I'm
> > sure it'd be possible to create some type of networking software in
> > assembly where each calc handles the incoming signal depending on where
> > it's coming from -- it'd require a lot of synchronization and thought,
> > obviously, and I don't know if you'd be able to have it recognize it
> > when a new calc joins without telling each calc explicitly, but it should
> > be able to be done to handle multiple calcs as long as the programming's
> > right, right?
>
> Or even neater, you could have the 'network' card
> microcontroller-based and let that tranlate between wahtever network
> protocol you'd liek to use and teh standard TI-85 routines... It'd be
> a little more expensive, and quit a bot harder to do, but it would
> mean that the network could also work on TI-82's as well...
> (You could add the capability to the card so that assembly programs
> could access the network directly or some such thing, so to increase
> upgradability..)
>
> As for a network protocol : you could use an ethernet-type protocol
> (i.e. when a calc wants to communicate with another calc, it sends
> out a message that all the othere calcs 'hear' which asks if the
> otehr calc is out there, somewhere, and if it is, it responds...)
> or we could use a TCP/IP type protocol and have a namesever, in which
> case, the transmitting calc looks up the other calc in the nameserver
> and then get's it no. and then tries to communicate with it
> directly..)
This whole thing is planned in the TINet, guys.
References: