I've looked into it briefly and here's what I understand so far. Correct me if I'm
wrong.
Apparently the main culprit is the H.323 protocol itself. I wanted to blame it on
Microsoft (for obvious reasons :) ), but Intel apparently shoved the protocol down
everyone's throat, and to compete everyone started using the protocol. Had it
undergone an RFC or other review process, probably someone would have spoken up and
said--"Um, this gets broken by just about every firewall there is." My guess is MS
made it worse with NetMeeting, but that's only based on the fact that they always do.
The protocol includes the client IP in the DATA portion of the packet. When the
firewall (STN) rewrites the packet HEADER with your external IP, it differs from
what's in the H.323 data in the packet which breaks something somewhere. Not sure, but
maybe the other side replies to what's in the data instead of the header?
What we'll need (and I believe someone in the LInux community is working on?) is a
proxy that's protocol aware, so that it changes both the data and header to the same
IP.
Or, the protocol itself could get fixed someday. Less likely.
BTW, I was able to get others to hear me, but I didn't get anything back.
pjb
At 12:52 PM 8/29/99 -0400, you wrote:
>It seems to me that sometime ago John was trying to figure out how to
>get Netmeeting to work through STN. Does anyone have it working? If
>not, what about CUSeeMe or something like it?
>
>Terry Horton
>
>
>
>
>_______________________________________________
>ShareTheNet maillist - [EMAIL PROTECTED]
>http://www.webserv.com/mailman/listinfo/sharethenet
>
>
Patrick Belliotti
PGP Key available at pgpkeys.mit.edu
PGP fingerprint: 705C B779 76B7 566F 78FC 6CC2 09F0 5EE6 C42D F0B7
_______________________________________________
ShareTheNet maillist - [EMAIL PROTECTED]
http://www.webserv.com/mailman/listinfo/sharethenet
