Re: network lircd modifications

[Jim P. <ji...@jt...>]

> The patch works by having both machines attach to a multicast group
> (using the -m command line option).  When a new message is recieved from
> the hardware a message is sent to the network.  When another machine
> hears this message, it forwards it to it's clients as if it had been
> received locally.  This works even if the receiving machine does not
> have a configuration for the specific remote.

I would argue that a much more useful way to do this would be to move
away from the named pipe and multicast scheme and modify LIRC to run a
TCP/IP server, like WinLIRC.  This way, you will:

1) have guaranteed packet delivery (you never get any acknowledgement
   for the multicast)
2) know how many clients are listening, and choose which ones get to
   hear what (solves the security issue)
3) be more compatible with networks that don't support multicast for
   whatever reason (switches, bridges, etc)
4) allow connections from arbitrary machines (for example, I may want
   a box of mine that is hosted at another site to respond to my
   signals)
5) not require LIRC or any other specific software at the receiving end
6) make it easier to implement sending over the network

What do others think?

(I'm not trying to say you had a bad idea; just offering what I think
is a better way to do it for the situtations that I can imagine)

-jim