Speak Freely and Reflector

[Steve Holmes]

Hey, I'm about to embark on setting up speak_freely on a linux box inside
an IP masqueraded network.  First question: should I be using the
speak_freely-7.2 as my base software? How much difference is there from
this one  and the CVS version? If I go with CVS, do I still need this base
7.2 version?  Is this stuff usable with the basic default OSS sound
support in linux or do I have to install ALSA?

Final question: I realize which ports to open up and forward through to my
inside machine but am a bit unclear with ipchains on which chains to
update.  I'm quite sure the forward chain needs to be set to accept the
UDP ports 2074:2076 and 4074:4076 but what about the input chain? Need I
accept there as well?  So far, I've pretty much just done IP masquerading
but haven't really organized my fire wall yet.

Thanks for any help on this two fold question.