keep-alive pings; sound-level meters

[sdsalsero]

Using TS 1.5, I am having continuing problems with my client losing the connection. I am connected to the Internet via a DSL line and a Linksys NAT router. I programmed the port-forwarding for 8765-6 (TCP and UDP) to my internal IP but it didn't help. I read in one of the forums that the server does a PING as a keep-alive check but my router does not respond to PINGs.

My suggestions:
- on initial connection, do a series of diags between the client and server to determine what ports and protocols are available
- the client should send a keep-alive packet rather than the server
- the client should better detect a lost connection, e.g. if no response from keep-alive

I would also like to see a sound-level meter for the client microphone. Many people have their mike set too high and their transmissions are distorted. They probably do this because their transmissions are not loud enough otherwise. So, you also need to have effective sound-level compression.

[Sensei Humor]

You can set the LinkSys to respond to ICMP (aka ping) by logging into the Web management page (point your browser to 192.168.1.1, assuming default settings) and clicking on the Advanced tab. At the bottom of the Filters page you'll find a radio button that says "Block WAN request". Set that to Disable and your router will respond to ping.

[sdsalsero]

Thanks Sensei, but I already knew that. I can also add my internal IP as the DMZ host, but that's not the point. I want Teamspeak to work with as few "requirements" as possible. If my client has established an outbound connection through the NAT, then that should be how the keep-alive is maintained.

[Sensei Humor]

I agree. I like blocking WAN requests, it's a security thing for me. I hope they resolve this in a future release.

[sdsalsero]

With TS v2.0 RC1, the problem seems to have gone away. It's possible we're using a different server config, though, so I can't say for certain that the new version fixed this. AND, since none of the developers have responded -- we'll never know if this suggestion was incorporated!

[Cstar_maxim]

TSv2, is a new beast compared to TSv1. It handles this in a differant manner. Yes TSv2 does not seem to have that problem.

The problem you stated was one of the many driving forces behind the generation of TSv2.